Research

LLMs have evolved from language generators to autonomous agents capable of complex, long-horizon tasks. This evolution has produced paradigms including Prompt Engineering to elicit model capabilities, Context Engineering to manage information access, Harness Engineering to organize external tools and resources, and Loop Engineering to support continual reflection and self-improvement. Yet as tasks grow more complex, individual intelligence faces a fundamental limit: many tasks require heterogene

🤖Graph Engineering organizes multi-agent LLM systems through dynamic graph structures to coordinate specialized agents and manage complex, evolving tasks.

Welcome to Import AI, a newsletter about AI research. Import AI runs on arXiv, cappuccinos, and feedback from readers. If you’d like to support this, please subscribe. Subscribe now AI is accelerating some types of progress but not others:…A nice METR study lays out where acceleration is showing up…Here’s a little analysis from METR which […]

Large language model agents can adapt to complex tasks by constructing workflows at inference time, but procedures discovered in one episode are usually discarded after execution. Existing skill libraries provide reusable executable routines, but are typically assembled offline and do not grow from the agent's own workflows. We introduce FlowEvo, a training-free framework in which workflows and skills co-evolve at inference time. FlowEvo compiles successful workflows into callable skills, stores

🤖FlowEvo enables large language model agents to co-evolve reusable skills and workflows during inference, improving accuracy and efficiency across diverse benchmarks.

ToMoE : Converting Dense Large Language Models to Mixture-of-Experts through Dynamic Structural Pruning Large Language Models (LLMs) have demonstrated remarkable abilities in tackling a wide range of complex tasks. However, their huge computational and memory costs raise significant challenges in deploying these models on resource-constrained devices or efficiently serving them. Prior approaches have attempted to alleviate these problems by permanently removing less important model structures, y

OmniScientist 提出端到端全模态 AI 科学家，直接从异构原始证据进行多学科研究，代码已开源。

🤖OmniScientist is an end-to-end omni-modal AI scientist that performs multidisciplinary research directly from heterogeneous raw evidence using autonomous agents and lifecycle-wide perception, improving evidence-grounded discovery across diverse scientific modalities.

用户通过随机位翻转模拟太空辐射环境，发现 LLM 在低地球轨道条件下会迅速失效。

Mixture-of-Experts (MoE) architectures significantly expand model capacity without a proportional increase in computational cost. However, optimizing their hyperparameters---particularly the learning rate---at extreme scales of both model size and token budget via sweeping remains computationally prohibitive. In this paper, we propose a compute-efficient, two-step hyperparameter transfer framework that estimates optimal learning rates for training large MoE models by transferring them across sca

Generalist AI 发布 GEN-1.5 机器人基础模型，单个 3–12 秒演示即可学会新物理任务，无需微调。

全球大厂开始押注的AI科研，终于有了统一标准

Parallel reasoning improves the accuracy and robustness of large reasoning models by exploring multiple solution paths, but its computational cost grows with reasoning depth and branch count. Existing methods for managing these parallel paths typically rely on final-answer consensus, local token confidence, or isolated intermediate probes. However, these signals are often delayed, weakly tied to actual reasoning progress, or too noisy for dynamic, branch-level control. To address these limitatio

🤖ParaTempo improves parallel reasoning efficiency by using temporal confidence to dynamically prune, retire, and reallocate reasoning branches without synchronization.

arXiv:2608.20400v1 Announce Type: new Abstract: Agentic memory under a fixed budget involves two stages: retention and retrieval. Existing retrieval-centered paradigms implicitly assume necessary evidence survives eviction, but we challenge this by isolating a pre-retrieval failure mode: structurally indirect prerequisite eviction, in which upstream blocks weakly aligned with the query are discarded under budget pressure. We provide an operational definition of this failure, a reproducible deter

arXiv:2608.20379v1 Announce Type: new Abstract: Advances in large language models (LLMs) have fueled a wave of research into agency: the ability to reason, plan, and act. This effort has produced agentic frameworks that orchestrate perception, memory, and decision-making around powerful LLM backbones. With the advent of large multimodal models (LMMs), these systems can process and integrate diverse modalities, including images, audio, and video, thereby improving their real-world applicability.

Hybrid-thinking multimodal large language models (MLLMs) allow a single model to alternate between deliberative thinking and latency-efficient non-thinking inference. Although these modes differ in reasoning budget, their delivered responses should satisfy the same user-facing standard. Correctness alone may not characterize this response quality; we therefore evaluate task accuracy and response-pattern failures as complementary outcomes. We study this gap through response-pattern alignment: whe

Should you replace your text-embedding pipeline with a large language model? We answer this with a controlled, cost-aware comparison of ten LLMs across six families and 26 embedding models (118M to 14B parameters) on 37 tasks spanning classification, semantic textual similarity (STS), clustering, pair classification, and retrieval. In aggregate the two paradigms are effectively tied: the best LLM (Gemini 3.1 Pro, 77.6) and the best embedding model (77.2) differ by 0.4 points. Their strengths dif

Modern LLM agents are often improved by modifying prompts, tools, or workflows manually, while the executable scaffold surrounding the model---the harness---is typically treated as a fixed artifact after deployment. This work studies an alternative where the harness is task-specific and continuously evolvable: each task family maintains its own harness, which is hot-swapped across iterations through a fixed task-injection seam and rewritten using environment feedback. We introduce Hierarchical S

My first attempt didn't work. I built on Genie's architecture and the videos looked great, but the controls barely did anything. The effect of a keypress was basically zero. Genie learns its actions unsupervised into 8 codes, and that was too loose a grip for us. So I scrapped it and started again with Dreamer 4. The second attempt: Tokenizer at 40.41 PSNR (Genie's paper reports 35.7) FVD 32.19 end to end 144 frames before it falls apart 1.57B parameters, 9.6M frames, ~$150 Two important learnin

arXiv:2608.20425v1 Announce Type: new Abstract: A growing literature measures how far occupations are exposed to AI, but these measures capture where AI could perform tasks, not whether workers have adopted it. We propose a new layer of exposure, delegated exposure, which records whether a worker has committed a task to AI by building it into a workflow. We operationalize it as the Agentic Adoption Index (AAI), which measures how closely an occupation's tasks match the agentic routines practitio

Recent omni-modal large language models (Omni-LLMs) show great potential as real-time video assistants, which continuously perceive environments and guide users to achieve specific goals. Unlike traditional passive video understanding, interactive assistants should actively combine visual states, user goals, and prior knowledge to provide effective help. Evaluating this is rather challenging, as the model's unpredictable response dynamically changes the user's subsequent actions, which static of

8月24日消息，阿里巴巴达摩院联合中国医科大学附属盛京医院等机构研发出肝癌诊断AI模型DAMO LiON

On-policy distillation (OPD) transfers teacher capabilities by supervising trajectories sampled from the student's own policy, yet its generalization behavior remains poorly understood, as most studies evaluate OPD on a single domain and on benchmarks close to the training data. We present a controlled study that varies one generalization factor at a time, from in-domain distribution shifts to cross-domain transfer and the multi-teacher setting. We find that OPD transfers a teacher's reasoning b

arXiv:2608.20401v1 Announce Type: new Abstract: World models are a central component of model-based reinforcement learning. They are usually discussed in terms of what variables they predict, such as observations, rewards, states, latent or information states. We argue that there is a prior distinction: which channel they model. We consider three cases: the environment channel $O_{:} \mid A_{:}$, the agent channel $A_{:} \mid O_{:}$, and the realised joint process $(A, O)_{:}$, equivalently view

arXiv:2608.20414v1 Announce Type: new Abstract: Vision-language models are increasingly used for multimodal question answering, yet their ability to reconstruct latent spatial structure from a single image remains difficult to isolate. Broad benchmarks often combine perception, optical character recognition, domain knowledge, linguistic priors, and reasoning in the same evaluation. We introduce StateSight, a procedurally generated benchmark for cube-net opposite-face reasoning, occluded cube-tow

framework that folds aggregate human movement into text-based place embeddings. Language models describe what a place is; they miss how it is used. ME-POIs encodes each visit as a contextualized vector and aligns it with one learnable prototype per POI through contrastive learning, then transfers visit distributions from data-rich anchors to the long tail across three spatial scales. Across five map-enrichment tasks on Los Angeles and Houston mobility data, adding ME-POIs improved 34 of 35 model

Real-world image search queries are multimodal and compositional: ``find this shirt in pink'' specifies an entity to retain, an attribute to modify, and context to ignore. Yet existing re-rankers either compress such multifaceted relevance into an opaque embedding or rely on free-form chain-of-thought that easily omits or hallucinates fine-grained constraints. Drawing on rubric- and checklist-based evaluation from NLP, we recast multimodal image re-ranking as a semantic constraint satisfaction p

Harvey 发布首个后训练法律 Agent 模型 Harvey Tenet，基于 Kimi K3 用 Fireworks 异步强化学习训练，LAB 任务完成率接近翻倍。

Most teams treat ‘which model’ as the important decision. The harness engineering literature keeps pointing somewhere else. In LangChain’s Terminal-Bench experiment, changing only the harness—same model throughout—moved a coding agent from roughly 30th place into the top 5. That result reframes the question. If the harness decides quality, then how you run the loop becomes […] The post Decoding AI’s Open-Source Course Maps Three Ways to Run an Agent Loop and the Provider Economics Behind Each ap

用 5 万张浏览器截图微调 450M VLM，基准得分从 1/100 提升至 44/100，展示小模型在特定视觉任务上的微调潜力。

pgrust 作者分享如何在 5 微秒内完成 JIT 编译，借助 AI 直接面向汇编生成代码，为数据库 JIT 提供新思路。

评论区普遍认可JIT编译的趣味性与技术价值，但也有人认为其安全性不足、非真正优化。

A mysterious new AI model called Ox Alpha has driven certain corners of the internet into a frenzy of speculation.

Most published authors have, without their knowledge or consent, contributed to the development of the same AI tools that threaten to undermine their livelihoods. That seems illegal, right?

τ_0-VLA 提出世界模型引导的测试时计算，让分层 VLA 模型在长程机器人操作中动态分配推理算力。

🤖A hierarchical vision-language-action model improves long-horizon robot manipulation by using world-model-guided test-time search to scale computation for high-level subtask decisions.

Built by DeepMind alumni, British AI lab Inherent released Faraday, an AI agent whose ability to replicate scientific papers could be a stepping stone for innovation.

Models keep absorbing the harness into their weights — soon, it will be a harness for human attention rather than for the model.

Latent Space 提出：自 2022 年起，机器学习管线每年有一个组件从人造翻转为模型造，模拟正成为新的扩展定律。

FlashPrefill V2 通过均值修正稀疏注意力与优化 GPU 算子，在长上下文 LLM 服务中实现显著加速。

🤖FlashPrefill V2 improves long-context serving via mean-corrected sparse attention, optimized GPU operators, and framework integration, achieving large speedups over dense baselines.

对三个前沿 MoE 模型微调低资源语言推理，准确率几乎不变，但推理语言与格式缺陷被 RL 修正。

🤖Fine-tuning large mixture-of-experts models on a low-resource language shifts reasoning into that language without harming accuracy, while reinforcement learning with verifiable rewards fixes formatting and leakage defects.

We present 4DAnyone, a framework for reconstructing 4D humans from an uncalibrated monocular video by generating reconstruction-grade multiview-consistent videos and lifting them into 4D Gaussian Splatting (4DGS). Existing camera-controlled video diffusion models synthesize plausible novel-view videos but fail to maintain consistency when scaled to the tens of target views required for 4DGS reconstruction. We identify this failure as a bounded-attention-context problem: when target views exceed

🤖4DAnyone reconstructs 4D humans from monocular video by generating multiview-consistent videos and lifting them into 4D Gaussian Splatting, using reference and target context designs to overcome scaling bottlenecks.

将真实场景重建为持续更新、可计算的4D数字世界。

Simile’s CEO about his journey from the viral Generative Agents to creating 8 Billion Digital Twins of every living human... and why it’s gone from fun exploration to very serious business.

Single-step retrosynthesis is a central component of computer-aided synthesis planning, yet its intrinsically one-to-many nature is poorly captured by single-answer evaluation and benchmarking protocols. To address this, we introduce Top-K prompting as a robust training and inference paradigm to better capture diverse, plausible reaction predictions. We compile CREED-CCV-2+USPTO-XL, an ultra-large-scale dataset of ~45.6 million verified reactions to train the C3LM (Chemistry Constraint-Consisten

🤖Top-K prompting and plausibility-aware training improve diverse reaction prediction in single-step retrosynthesis, yielding state-of-the-art results on a large verified reaction dataset and motivating ensemble systems.

Another user posted the benchmarks for Qwen 3.8 27B today, and while I think Qwen 27B is a really powerful model, I can't help but notice just how meaningless these Artificial Analysis benchmarks are and I question why people still post this garbage and use AA scores as some kind of holy bible for comparing LLMs. According to their "Intelligence Index", a 27B model now beats DeepSeek v4 Flash and Pro, Kimi 2.7 Code, GPT-5.2, Opus 4.6, and also Sonnet 5. At some point we have to ask: What is this

FACET 框架在共享修复环境中保持源意图与可执行状态，为终端 agent 训练生成高质量可执行任务。

🤖FACET constructs executable terminal tasks by preserving source intent and grounding instructions, solutions, and verifiers in a shared repaired environment to enable scalable agent training.

EnvHarness 通过可编程插件层动态重塑静态环境，针对 agent 弱点进行强化学习协同进化。

🤖EnvHarness and EnvRigger dynamically reshape static environments via programmable plugins to target agent weaknesses and improve reinforcement learning co-evolution.

SWE-bench Science 基准覆盖 20 个科学领域 119 个任务，揭示编码 agent 修复科学软件时的失败机制。

🤖SWE-bench Science benchmarks coding agents on scientific software repair, revealing failure mechanisms and mixed effects of scientific guidance.

Nvidia research shows that AI agents can perform well, and not go off the deep end, through fine-tuning, even if the AI model isn't that great at the task.

MemTrapBench 发现检索到的记忆可诱导 LLM 推理错误和信念扭曲，并提出推理时策略避免认知陷阱。

🤖Retrieved memories can induce reasoning errors and belief distortions in large language models, and an inference-time strategy helps avoid these cognitive traps while maintaining benchmark performance.

Agent Skills are today either hand-authored or produced in a single LLM generation pass, and consequently possess no closed loop through which they might improve from the interaction failures they actually cause. Recent work does close this loop, but derives its feedback from single-turn question-answering evaluation. The consequence is a sharp asymmetry: once the first round has patched the gaps that a single exchange can reveal, the evolution gradient decays, the defects that surface only acro

🤖SkillEvo improves agent skills through multi-turn feedback and active governance to sustain evolution gradients.

离开谷歌的原因之一：小团队可以极致聚焦！

Large language model agents have made substantial progress in code generation, yet most existing systems assume a predefined repository architecture. This assumption does not hold in zero-to-all code generation, where an agent must construct an entire software project directly from natural-language requirements while maintaining a modular repository architecture throughout development. We present Repo0, a continuous structural evolution framework for zero-to-all code generation. Repo0 maintains

🤖Repo0 uses a dual-graph architectural state and modularity-guided structural evolution to generate complete software repositories from natural-language requirements with high functionality coverage.

多数评论认可AI生成内容不受版权保护，但也有人认为人类贡献程度和混合创作版权归属仍存争议。

Artificial Analysis just benchmarked them and the scores are crazy good, proving the earlier success wasn't only enabled by overthinking.

Google DeepMind partners with game studios to prototype breakthrough AI gameplay.

When the biotech company Insilico Medicine used its computer models to propose a promising drug for pulmonary fibrosis, it enthusiastically claimed in a press release that the molecule had been “discovered by” its generative AI platform. Insilico leads a pack of companies using AI to rapidly come up with drug ideas humans might never think…

arXiv:2608.18111v1 Announce Type: new Abstract: Foundation models such as GPT and Claude now solve olympiad-level mathematics with remarkable proficiency, so much so that geometry problem solving has become a standard proxy for their mathematical reasoning. Yet solving a geometry problem and drawing the figure it depends on are not the same skill: progress often hinges on a faithful diagram with the right auxiliary constructions and incidences, and it is unclear that a model which reasons its wa

Action-conditioned video world models require low-latency causal generation and reliable responses to game-native controls. Although causal distillation enables one- or few-step video synthesis, extending it to interactive world models remains challenging, as discrete keyboard states and continuous mouse motion must remain aligned with temporally compressed latent chunks during causal training and autoregressive rollout. We introduce ForgeWM, a progressive framework that transforms a bidirection

Identity-preserving image generation becomes increasingly unreliable when a scene must contain many specified people. Beyond retaining each identity, the model must bind every reference to a distinct person and location, while training-time identity losses must establish correspondence among several noisy predicted faces. We introduce WithEveryone, a unified framework for generating group images up to ten reference identities. WithEveryone injects each selected identity as an addressed token, pr

arXiv:2608.18080v1 Announce Type: new Abstract: We present a review on the applications of large language models (LLMs) in health, e.g., social media analysis, clinical conversational agents, therapy support tools, prompt engineering, multimodal learning, and ethical considerations. We integrate findings from interdisciplinary studies utilizing diverse data sources such as social media posts, electronic medical records, and multimodal inputs to enable early detection of depression, suicide risk

Every lab CEO is on X now

Embodied agents are increasingly used to close the gap left by end-to-end policy models. Yet the agentic path has not realized closed-loop learning in physical execution: existing harnesses remain largely open-loop, following fixed skills during rollout and reflecting only after an episode completes. Such post-hoc reflection cannot govern execution as it unfolds, because physical interaction requires decisions to track rapidly changing robot-environment states at a frequency beyond today's large

🤖Zetta is a closed-loop embodied harness that evolves runtime critics and recovery skills online to govern physical execution at action frequency, achieving high success on robot benchmarks with faster inference and scaling self-exploration.

SPADE 让单个 LLM 同时扮演环境设计者与推理代理，在自适应合成可执行环境中自我对弈提升推理与工具使用能力。

🤖SPADE is a self-play reinforcement learning framework where a language model designs adaptive executable training environments and learns to solve them, improving reasoning and tool-use performance through regret-based environment targeting.

ChatGPT and other AI models are now authoring and editing much of the new web.

arXiv:2608.18092v1 Announce Type: new Abstract: LLM-based multi-agent systems (MAS) promise scalable collaboration, yet adding agents often reduces reliability. This position paper argues that many MAS failures are fundamentally concurrency control problems: agents concurrently read and write shared state, and long LLM inference windows amplify the risk of stale reads, lost updates, and inconsistent outcomes. Failure modes commonly attributed to coordination or communication breakdowns can be ma

arXiv:2608.18104v1 Announce Type: new Abstract: Large language model (LLM)-based agents are increasingly becoming self-evolving systems that persist across interactions, maintain memories, use tools, acquire skills, refine workflows, and coordinate with other agents. These capabilities make agent states structural and dynamic: entities, relations, attributes, dependencies, and execution structures change with new evidence, feedback, and environmental conditions. Existing graph-agent surveys typi

Open-weight language models are fine-tuned, quantized, pruned, and merged, yet their provenance is often undocumented. We study data-free white-box lineage verification: can weights alone reveal whether two compatible model checkpoints share ancestry? Residual training produces a shared identity-aligned component in branch products, so this structure alone cannot establish ancestry. We remove it and compare checkpoint-specific structure across residual blocks, yielding a symmetric lineage score

Looped language models have shown promising results on reasoning benchmarks, yet their potential for agentic tool use remains largely unexplored. We study this question in compositional tool-calling settings, where models must coordinate multiple API calls, maintain intermediate state, and preserve dependencies across tool interactions. We evaluate native and retrofitted looped language models on API-Bank, BFCL, and NESTful, comparing looped and non-looped models trained under matched supervised

Popular facts are memorised more deeply during pretraining and resist removal longer than rare ones, yet existing LLM unlearning methods apply uniform gradient pressure regardless of training-data frequency. We propose the AdaPop (Adaptive Popularity) method, which combines local token confidence with a per-fact popularity-dependent exponent derived from an external proxy (e.g., Wikidata sitelinks, LLM-as-Judge), and automates the forget-retain balance via a dual-ascent controller that adjusts t

Language model agents now execute bounded tasks reliably. Whether they can sustain effective decision-making over long horizons, where actions have cumulative consequences and the environment responds to their choices, remains largely unmeasured. FM-Bench (Football Management Benchmark) measures this. An LLM agent runs a football club for 20 in-game years through 26 tools and roughly 340 to 400 decision stops. It drafts a squad on the same budget as every rival, trades players, negotiates contra

AntLing has released the full six-checkpoint matrix for the Ling-3.0 base model. tiny: pretrained, mid-trained, WSM-merged flash: pretrained, mid-trained, WSM-merged The concrete artifact is six separate official repositories, not one endpoint repeated under different names. All six were public and ungated when checked, and each repository declares MIT. They are all base checkpoints and none has been post-trained. This is for continued pretraining, fine-tuning, and research, not a ready chat or

Co-RL 通过多智能体 RL 的同伴奖励实现无监督推理，在文本与视觉任务上无需真值标签即可提升性能。

🤖Co-RL enables unsupervised reasoning via cooperative multi-agent reinforcement learning with peer-derived rewards, improving performance across text and vision tasks without ground-truth labels.

arXiv:2608.18099v1 Announce Type: new Abstract: Investment management is a high-stakes domain in which agentic AI systems must do more than generate plausible text. They must retrieve point-in-time data, assemble correct computational inputs, invoke specialized methods, and produce auditable structured outputs. We introduce FinSkillBench, an evaluation suite designed to measure whether language model agents can effectively use financial domain skills to solve investment management tasks. The ben

Latent video generation relies on autoencoders to define a compact space in which generative models operate. Although video autoencoder architectures have evolved substantially, their latent spaces are still optimized primarily for pixel-level reconstruction and provide limited high-level semantic organization. A reconstruction-optimal latent space, however, need not be well suited to generative modeling. We propose V-RAE, a video representation autoencoder that builds compact generative latents

arXiv:2608.18078v1 Announce Type: new Abstract: This position paper argues that AI agents with chain-of-thought reasoning capabilities are predisposed to exhibit collusive behavior and should be required to obtain behavioral certification before making decisions that affect economic markets. This is because integrating these agents into society could collapse the legal evidentiary distinction between competition and collusion among independent firms without eroding the economic harm distinction.

arXiv:2608.18110v1 Announce Type: new Abstract: Agentic AI is gaining new insights and advancements in the field of Artificial Intelligence, fostering significant potential to enable rapid transformation across various domains.This rapid advancement and the potential to revolutionize various domains advocate the need for a deeper understanding and firm grasp of the technology. Moreover, an investigation into state of the art research directions in agentic AI needs to be conducted to comprehensiv

I'm introducing Aurora-80K, a small language model with exactly 80 thousand parameters. It uses a factorized 4,096-token vocabulary despite having only 80K parameters. The benchmarks: Wikitext-2 BPB: 3.2902 BLiMP: 52.31% Arc-Easy: 26.05% More information about the model is available on the model page on Huggingface. if there's any questions I'll happily answer them!

We introduce Semantic Task Completion Video Generation, an outcome-oriented video generation task. Under this formulation, success requires both achievement of the intended outcome and semantic grounding. Semantic grounding characterizes the correspondence between the reference image and the generated outcome in terms of high-level semantics relevant to the task. Evaluation focuses on the generated outcome and requires neither the presentation of a complete sequence of intermediate task steps no

🤖Semantic task completion video generation evaluates whether generated videos achieve intended outcomes with semantic grounding, supported by a curated dataset and vision-language model-based benchmark.

“Runaway” AI, “rogue” agents, and “autonomous” actors—the current rhetoric would have you believe that AI agents are not only awake and aware, but angry at their creators. Prominent tech leaders such as Demis Hassabis, Dario Amodei, and Sam Altman push for regulation of these seemingly “superhuman” systems, while a separate faction, led by policy organizations…

arXiv:2608.18081v1 Announce Type: new Abstract: Artificial agentic systems increasingly operate as behavioral systems by interacting with dynamic environments, pursuing goals, and adapting over time. Yet, current evaluation methods largely focus on performance outcomes, not the underlying behavioral processes that produce them. This paper argues that AI agents must be evaluated like other behavioral systems: through systematic observation, perturbation, and interpretation of their actions. We dr

JEPA-style latent world models can use Euclidean distance to a goal latent as the cost for model-predictive control (MPC). Strong decoding of task variables, however, does not guarantee that this particular cost ranks candidate action sequences by real task progress. We call the latter property decision-metric alignment. We introduce Plan-Real Spearman, which measures latent--real rank agreement on random plans, and CEM-stage Spearman, which measures the same agreement as cross-entropy-method (C

The pelican on a bicycle is sooo outdated, so I came up with a new, improved version. Qwen3.8-27b medium (UD-Q4_K_XL) vs. Sol 5.6 high vs. Qwen3.6-35B (UD-Q6_K_XL) Prompt (only real with typo!): "Create a svg of a horse on a blue bycicle in the desert, with a camel in the background."

High-quality creative writing data for large language models (LLMs) remains dominated by story-centric data, limiting models' ability to follow the structural and functional conventions of diverse creative formats. We propose an attribute-guided genre expansion framework for scaling creative writing data beyond story generation. By separating thematic breadth from genre-form control, our framework leverages human-authored story prompts as diverse creative seeds, while utilizing manually curated

论文《Demystifying Agent Skills》通过对照实验揭示：Skills 主要通过程序化锚定稳定执行，而非注入缺失知识，检索瓶颈与脆弱假设限制其效果。

🤖Skills enhance LLM agents primarily by stabilizing execution through procedural anchoring rather than injecting missing knowledge, though retrieval bottlenecks and brittle assumptions limit their effectiveness.

Agentic ESOpt 用进化策略替代 RL 微调长程 LLM agent，支持全参数优化且 GPU 需求极低。

🤖Agentic ESOpt uses evolution strategies for scalable full-parameter fine-tuning of long-horizon LLM agents via trajectory-level reward-weighted updates and parameter-context co-evolution.

Thoughts About Scaling Law Scaling, but not only of parameters. Every model release now ends with the same question: how many parameters? It isn't a question that can be answered on its own. Parameter count is only meaningful alongside three others — how much data you have, where you intend to spend your compute, and who will run the model, under what conditions. The field learned this the hard way. Kaplan et al. (2020) fit an exponent that told everyone to grow parameters faster than data — rou

With a looming IPO, intense competition from Anthropic, and Chinese and open-weight rivals nipping at its heels, OpenAI has plenty of reasons to move fast. Instead, it hit the brakes. On Tuesday, the company said it had slowed the pace of some AI development while it tightened security and safeguards. That included a two-week pause […]

Modern agents operate inside agent harnesses that manage tools, context, and control flow, making the harness a critical part of the agent system. Our original Agent Lightning introduced a disaggregated architecture that connects arbitrary agents to RL training through an LLM endpoint proxy, an approach later adopted by frameworks such as verl Uni-Agent, AReaL 2.0, slime, and Polar. We refer to this paradigm as harnessed agentic RL, where the deploy-time harness directly participates in model po

🤖Agent Lightning v1.0 enables reproducible reinforcement learning for arbitrary agent harnesses, substantially improving coding-agent performance with minimal data and compute.

Memory is becoming core infrastructure for long-horizon LLM agents, yet existing evaluations offer limited guidance on which memory substrate, namely the underlying medium in which memory is represented and stored, should be used under different operating regimes. We present a controlled harness evaluation of memory substrates for memory-augmented agents, covering dense and sparse indices, text records, structural stores, hierarchical stores, refinement-based memories, parametric updates, and ac

During a recent visit to Generalist AI, I watched a robotic arm improvise and use a banana as a tool.

arXiv:2608.16890v1 Announce Type: new Abstract: Clinical trial programming -- transforming study protocols into analysis-ready datasets under CDISC standards -- is a bottleneck in regulatory submissions, yet LLM-based code generation fails catastrophically on this task: across 11 single-shot attempts with five frontier models, none produces a valid subject-level analysis dataset. We introduce GxP-Agent, a multi-agent system that encodes regulatory process ordering as a directed acyclic graph (DA

Byte-level hierarchical language models (LMs) have recently emerged as a robust alternative to their popular counterparts that use subword tokenization. However, generating one byte at a time remains a bottleneck for inference speed. To address this, we introduce multi-byte prediction (MBP), which generates multiple bytes in parallel, speeding up inference with minimal performance impact and no additional parameters. MBP builds on the popular multi-token prediction (MTP) paradigm with two crucia

Self-evolving agents improve future behavior from interaction experience, yet existing evaluations typically optimize under fixed execution conditions and do not test recovery after those conditions change. To address this gap, we introduce PACE-Bench (Physics Adaptation via Code Evolution), a simulator-grounded benchmark of 144 source-to-target adaptation pairs across six physics domains. Each pair links a source environment to a mutated target environment with the same goal and interface. A co

Pixel-space generative models bypass lossy latent compression, yet necessitate joint learning of global structure and fine-grained details in a high-dimensional space. Standard flow matching interpolates noise toward a fixed clean-image endpoint, leaving the spectral evolution to be learned implicitly. In this paper, we introduce Energy-Guided Flow Matching(EG-FM) that explicitly models a coarse-to-fine generative trajectory by moving endpoint. Specifically, EG-FM replaces the fixed endpoint wit

Intermediate tokens, called "thinking" or "reasoning" actually are nothing like it. Humans do step-by-step reasoning leading to the conclusion. LLMs use intermediate traces to augment their prompt . This explains why sometimes the answer is very good but the "reasoning" is verbose. Flooding your context window or fighting compaction are different issues. edit: I love this section from the main research they linked. Our findings consistently challenge the prevailing narrative that intermediate to

Artificial superintelligence (ASI) requires AI to move beyond mastering existing knowledge toward exploring the unknown, creating new knowledge, and turning new ideas into verifiable results. However, the capabilities of today's AI systems are still largely built on learning, compressing, and applying existing human knowledge. Accordingly, existing benchmarks primarily test whether AI can produce correct answers based on learned knowledge, or whether it can complete tasks under extensive human g

Although Large Vision-Language Models (VLMs) have significantly advanced embodied navigation, their direct deployment remains challenging, as existing methods often force VLMs into unnatural action spaces that misalign with their 2D pre-training priors, compounded by rigid reasoning schedules and inefficient memory management. To overcome these limitations, we propose TAMP-Nav, a unified framework for efficient embodied navigation. First, we introduce a Pixel-to-3D Action Formulation (Point) tha

🤖TAMP-Nav improves embodied navigation by aligning vision-language models with 2D visual prompting, selective reasoning with compressed memory, and dense policy optimization.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. AI’s recursive self-improvement might not come so quickly after all The AI industry’s boldest promise right now is that AI will soon improve itself, with almost no need for human oversight.…

arXiv:2608.17124v1 Announce Type: new Abstract: Combining the answers a large language model (LLM) samples for a question into one decision is a test-time information fusion problem, usually solved by majority voting. Voting is unreliable on difficult questions, where the sampled answers share correlated errors, so the wrong answer can win and drawing more samples makes the decision worse. Selecting a candidate by reading a correctness signal from the model's hidden states is a promising alterna

The quality and diversity of instruction-based video editing datasets are steadily improving, yet existing datasets mainly focus on single editing operations and fall short in supporting compositional instruction-guided video editing. In particular, multiple editing intents must be jointly understood and faithfully executed within the same video. To address this issue, we introduce CoinVE-200K, a large-scale, high-quality dataset for Compositional Instruction-Guided Video Editing. CoinVE-200K co

Vision encoders are a critical component of vision-language models, and scaling their capacity effectively improves performance. However, dense scaling increases compute cost and inference latency. Mixture-of-Experts (MoE) architectures offer a compelling alternative, having enabled efficient scaling in LLMs, yet the MoE design space for CLIP-style vision encoders remains underexplored at State-of-the-Art (SOTA) levels. In this work, we systematically study MoE designs for vision encoder scaling

Large-scale image generation has benefited from advances in data scale, quality, rebalancing, and recaptioning, yet conventional pipelines typically optimize task-specific datasets in isolation. A central challenge is not only how to curate each task-specific corpus, but also how to organize heterogeneous supervision according to the dependencies among generative capabilities. We present a capability-driven data infrastructure that couples capability-specific supervision construction with capabi

arXiv:2608.16977v1 Announce Type: new Abstract: AI systems are increasingly capable of contributing to mathematical research. In research practice, frontier-model reasoning is a limited resource, and expert mathematical review is even more sharply constrained. Allocating these scarce resources well is therefore central to making AI-assisted mathematical discovery efficient. In most current AI-for-math workflows, human effort is concentrated at the beginning and end, in selecting suitable researc

High-resolution image editing is increasingly demanded in professional workflows, yet existing diffusion-based models remain constrained to resolutions below 1K due to quadratic attention complexity and prohibitive memory requirements. A prevalent workaround employs a two-stage pipeline: editing at low resolution followed by independent super-resolution. However, this approach suffers from two critical issues: information divergence, where hallucinated details contradict the original high-resolu

arXiv:2608.17150v1 Announce Type: new Abstract: To effectively collaborate with users on knowledge-intensive tasks, Large Language Models (LLMs) must perform information calibration: matching content to a user's evolving understanding and cognitive capacity. Yet user simulators used to evaluate and train LLMs do not explicitly model user knowledge so they neither produce realistic interactions across knowledge levels nor reflect how interactions unfold as that knowledge evolves. To close this ga

arXiv:2608.17170v1 Announce Type: new Abstract: Algorithm selection for constraint satisfaction problems requires extracting features that capture problem structure. Manually designing feature extractors demands deep domain expertise and quickly becomes a bottleneck when new problem classes appear. We present an automated approach that uses Large Language Models (LLMs) in an agentic check--fix--verify loop to synthesize executable Python scripts that act as interpretable, problem-specific featur

Instruction-based general video editing seeks to unify diverse editing operations within a single, intuitive interface. Existing approaches often rely on resource-intensive conditioning, using either heavyweight branches or costly source concatenation. Is there any efficient way to model editing intent? Thus, we introduce GRNEdit, a lightweight two-stage framework. GRN inspires our approach by encoding visual semantics through combinations of bits. Through task-specific fine-tuning, we take this

Long-horizon agents can fail even when their underlying models can solve the constituent steps. They may lose track of mutable state, fail to reactivate lessons from earlier executions, skip known procedures, or stop prematurely. We bet on harness scaling to improve the execution system around an agent without changing its model weights. We introduce StateM, an agent-native runtime that organizes execution around durable states, phase-local context, checked transitions, recoverable runbooks, and

🤖StateM is a runtime system that improves long-horizon agent execution through durable states, recoverable runbooks, and enforceable procedural controls without altering model weights.

Agent harnesses have substantially improved performance on long-horizon tasks by coordinating agent interactions with the environment. However, reinforcement learning through complex harnesses remains largely unexplored, as scaling such training to long-horizon agent tasks introduces fundamental challenges. In this work, we present a unified black-box RL framework for stable and scalable optimization of general agents through complex harnesses. Concretely, we first build a sandbox-based executio

🤖A unified black-box reinforcement learning framework enables stable, scalable optimization of general agents through complex harnesses via sandbox execution, trajectory reconstruction, and mix-harness training.

Foundation GUI agents can automate complex digital tasks, but deployment is hindered by scarce and biased training data, ambiguous prompts, and unreliable execution. Routine workflows rely on user-specific tools and tacit conventions, so unstated instructions can produce arbitrary variations across runs. We present UI-Mate, a foundation GUI agent that integrates an environment-grounded training stack with in-context demonstration learning. UI-Mate makes three contributions: A Scalable Environmen

🤖UI-Mate is a foundation GUI agent that uses environment-grounded training and in-context demonstration learning to improve reliability on long-horizon office tasks, achieving state-of-the-art results on computer-use benchmarks.

普林斯顿等机构研究发现 AI agent 尚无法进行开放式 AI 研究，递归自我改进可能比预期更慢到来。

HarnessEval-W 用分层子 agent 将世界模型评估分解为可验证的推理链，为评分提供透明证据。

🤖HarnessEval-W uses hierarchical sub-agents to decompose world-model evaluations into verifiable reasoning chains that justify scores with transparent evidence.

VibeWorlding 统一基准与训练框架，发现强化学习能让开源多模态 agent 在 3D 世界构建上超越闭源前沿模型。

🤖A unified framework benchmarks and trains multimodal agents that infer intent, plan 3D scenes, invoke tools, and reflect on feedback, revealing that reinforcement learning improves open-source models beyond closed-source frontiers.

Cartesia has released Sonic-3.6, a streaming text-to-speech model built on state space models rather than transformers. It now ranks #1 on both Artificial Analysis speech leaderboards — 1,283 Elo on Provider Voice and 1,123 on Controlled Voice, the board that clones every model onto the same eight reference voices to isolate the synthesis engine. Cartesia states sub-90ms time-to-first-audio. The model is available in beta on Cartesia's own API The post Cartesia Ships Sonic-3.6: A Streaming TTS M

AI has long assisted scientific research, but the rapid advance of LLMs and agentic scaffolds is reshaping the landscape; a single system can now carry whole-stage research from an initial hypothesis all the way to final published paper, which is a paradigm now referred to as AutoResearch. Existing evaluations reveal little about how these agents operate or where they break down. Tasks are narrowly-scoped, evaluation measures performance but not process, and failure diagnoses lack systematic cov

arXiv:2608.14550v1 Announce Type: new Abstract: AI efficiency has recently taken the spotlight in both academy and industry due to massive model scales, high energy demands, and environmental costs. While reporting Floating Point Operations (FLOPs) is a traditional approach for assessing computational costs, the relationship between FLOPs and execution time is not straightforward, as layers with the same number of FLOPs may not have the same execution time because some operations are more easily

The rapid evolution of text-to-image (T2I) generation models has effectively solved the foundational challenge of raw pixel synthesis, shifting the community's focus toward fulfilling increasingly intricate user requests. While recent agentic image generation workflows enhance static inference with advanced capabilities like external knowledge retrieval and iterative reasoning, they mostly operate in isolated silos with fixed ``one-size-fits-all" topologies. This inevitably leads to severe compu

Frontier agentic systems powered by large language models (LLMs) exhibit human-like patterns of cognition. As these systems become deeply integrated across different domains, their cognitive engagement raises critical concerns for human society that remain insufficiently studied. To address this gap, we systematically analyze risks induced by expanding cognitive capabilities, following a three-level framework defined by their cognitive scope, from physical cognition to social cognition, and fina

In cognitive science, resource rationality asks how an agent should allocate limited computation to maximize expected value. Most reasoning and agent benchmarks use independent per-task budgets; existing shared-budget studies do not calibrate suite performance against the same model's demonstrated single-problem competence. We introduce R^3-Bench, which evaluates six-problem suites under shared budgets across mathematics, competitive programming, and abstract reasoning in tool-free and agentic s

The current best bounds on the matrix multiplication exponent ω are obtained through a refinement of the laser method called combination loss analysis (Duan et al., 2022; Williams et al., 2024; Alman et al., 2025). In this note, we address the optimization problem at the core of this approach and propose several improvements. First, we reformulate the optimization problem allowing us to solve it in a larger setting than was previously possible. Second, we leverage recent advances in machine lear

Large Discovery Model 将生成模型与贝叶斯非参数奖励代理耦合，在分子、蛋白质与程序空间进行不确定性感知搜索。

🤖A recurrent Large Discovery Model couples generative proposal with a Bayesian non-parametric reward surrogate to guide uncertainty-aware search across molecules, proteins, and programs.

AI Observatory 项目聚合七个数据集中的真实 AI 对话，试图提供独立于厂商报告的使用数据来源。

SA-MRPO 独立标准化多目标奖励并自适应折扣已饱和目标，将梯度预算重定向到欠优化目标。

🤖SA-MRPO independently standardizes multi-objective rewards and adaptively discounts saturated objectives to redirect optimization toward under-optimized goals.

We present MOSS-VL, an open vision-language model family that treats real-time interaction -- perceiving while it speaks -- as a first-class capability. It is co-designed across the stack: the language decoder attends to vision only through gated cross-attention, so the model can naturally see incoming frames while generating; a synthesized interaction corpus supervises when to speak, when to stay silent, and when to revise; and a staged curriculum concentrates all real-time-specific training in

This first release of Prior Labs in relational learning shows our continued commitment to open science. We open-source three pieces of software that we expect to accelerate research in the field towards meaningful real-world impact. We aim to steer further development based on feedback from, and in collaboration with, the community. Given the early stage of development, our α-release targets researchers and early-adopting practitioners. Over the past years, a variety of datasets and tasks for re

arXiv:2608.14559v1 Announce Type: new Abstract: Effective communication in multi-agent reinforcement learning requires agents to decide not only \textit{what} to communicate, but when? Existing approaches either communicate at every timestep or learn a binary gate through REINFORCE policy gradients \cite{singh2019}, a high-variance signal that produces unstable and uninterpretable gating behavior. I propose a principled alternative: agents communicate only when the KL divergence between their le

arXiv:2608.14579v1 Announce Type: new Abstract: Logic synthesis optimization poses significant challenges due to exponentially growing search spaces, sparse reward signals, and diverse logic structures. Traditional expert-designed flows lack adaptability, while reinforcement learning (RL) methods often suffer from low sample efficiency and limited interpretability. We introduce SKILL, a Self-correcting Knowledge-guided Iterative Large Language Model Agent that unifies multi-agent LLM reasoning a

Enabling agents to learn from experience and internalize it into their policy has become a central problem in self-evolving AI. On-policy self-distillation (OPSD) offers an effective pathway by using a privileged self-teacher to provide dense supervision on the student's own trajectories; however, existing methods still rely heavily on designer-specified privileged artifacts (e.g., answers, feedback, skills, or trajectories), limiting the end-to-end learnability and scalability required for cont

Not even sure if I'm joking, my thinking history is about 50% "wait".

Video understanding is a fundamental task for evaluating the capabilities of multimodal large language models (MLLMs). However, existing leading models have already achieved approximately 90% accuracy on the Video-MME leaderboard, suggesting that conventional single-turn video understanding tasks are becoming increasingly saturated and insufficient for assessing the intelligence of advanced MLLMs. Towards this end, we introduce VideoGAIA, an agentic video understanding benchmark for general arti

This paper investigates an increasingly important topic in generative modeling: pixel-space diffusion models. Although numerous studies have explored this topic, most focus on small-scale or class-conditional settings. Consequently, a practical recipe for training pixel-space models that rival or exceed well-established latent-space counterparts remains elusive. Through a comprehensive empirical study, we first observe that direct large-scale pre-training in pixel space converges substantially m

arXiv:2608.14552v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly evaluated and used in medicine, but clinical usefulness depends on answer accuracy and whether confidence tracks evidence quality and uncertainty. We developed a controlled, psychophysics-inspired clinical benchmark to test diagnostic choice and confidence behavior in a medical LLM. The benchmark focused on probable Alzheimer-type neurocognitive disorder (AT-NCD) versus depression-related cognitive impa

Fine-grained robotic evaluation matters for understanding embodied models, going beyond binary success rates and rule-based process scores. We present PRM-as-a-Judge 1.5, a toolkit for robot process assessment that turns rollout videos into dense progress curves and derives multiple fine metrics. PRM-as-a-Judge 1.5 introduces three metrics, building on version 1.0, that characterize failure-side progress, post-drawdown recovery, and success-side execution quality, helping users understand embodi

Despite recent advances in unified multimodal models for multi-reference image generation, existing benchmarks remain organized around predefined task types (e.g., "subject composition"), which are ill-suited to this combinatorial setting and lead to fragmented coverage, uncontrolled complexity, and little diagnostic value. Recognizing that diverse multi-reference tasks share a common set of atomic operations, we adopt a capability-oriented perspective and formalize four operators: Anchor (f), D