-
1. Learn2Play Bench: How Well Do LLM Agents Learn from Experience in Unfamiliar Environments?
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
91 HF upvotes · 1 comments
来自 Hugging Face Daily Papers,主题偏「Agent、Reasoning、Eval/Data」。摘要显示它主要讨论 Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability is therefore important for understanding how effectively agents acquire and use new knowledge. Existing benchmarks have sought to eval... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果你要找可复现 demo、开源工具或产品化线索,建议点开项目/GitHub;否则先看中文摘要即可。
AgentReasoningEval/DataAI InfraRobotics
Original abstract
Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability is therefore important for understanding how effectively agents acquire and use new knowledge. Existing benchmarks have sought to evaluate this ability, but they primarily evaluate tasks whose rules are provided in the instructions or already familiar to pretrained models, making it difficult to distinguish learning from interactions from reasoning with existing knowledge. To address this, we introduce Learn2Play Bench, a benchmark of newly designed text-based games, whose rules are novel or counterintuitive, requiring agents to acquire knowledge through interaction rather than rely solely on pretrained knowledge. These games provide reproducible feedback and automatic scoring, enabling controlled evaluation of learning across repeated attempts. We also vary game instances to test whether agents can apply what they have learned to new situations. Therefore, we evaluate how backbone models, self-evolving methods, and agent harnesses affect agents' learning ability, revealing three findings: (1) Experience retention: Retaining complete records of actions and feedback can support more effective learning than summarizing these experiences into rules or strategies. (2) Human agent gap: Top-performing human players achieve higher peak scores than the evaluated...
-
2. SuperNav: An Agentic Navigation System for Any Task in Any Scene
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
58 HF upvotes · 1 comments · 10 GitHub stars
来自 Hugging Face Daily Papers,主题偏「Agent、RAG/Memory、Multimodal」。摘要显示它主要讨论 General-purpose service robots need navigation systems that can handle diverse human requests in unfamiliar environments, combining task generality with scene generality. Some existing methods fine-tune multimodal large language models (MLLMs) to predict navig... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果你要找可复现 demo、开源工具或产品化线索,建议点开项目/GitHub;否则先看中文摘要即可。
AgentRAG/MemoryMultimodalEval/DataAI Infra
Original abstract
General-purpose service robots need navigation systems that can handle diverse human requests in unfamiliar environments, combining task generality with scene generality. Some existing methods fine-tune multimodal large language models (MLLMs) to predict navigation actions, making their behavior dependent on the coverage of navigation training data and potentially limiting generalization to new requests and environments. Our key insight is to let the MLLM focus on interpreting requests, understanding scenes, and making decisions while preserving its general-purpose capabilities and delegating motion execution to navigation tools. To realize this idea, we introduce SuperNav, which equips a pretrained MLLM with a specialized agent harness without navigation-specific fine-tuning of the MLLM. Our harness supports these decisions with Navigation Skills, agent-oriented Tools for physical interaction, and task-progress and context management. A unified visual-point interface connects decision-making to motion by allowing the model to specify destinations directly in images and revise its decisions from execution feedback. Together, these components support sustained navigation across different task requirements and environments. SuperNav outperforms four evaluated baselines on instance-level, multi-object, and demand-driven tasks. Category-level evaluation on HM3D and deployment on a...
-
3. Multi-Agent Egocentric World Model with Fine-Grained Embodied Interaction
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
36 HF upvotes · 1 comments · 14 GitHub stars
来自 Hugging Face Daily Papers,主题偏「Agent、RAG/Memory、Multimodal」。摘要显示它主要讨论 Egocentric world models predict first-person observations conditioned on an agent's actions, but most focus on a single agent. Real embodied settings often involve multiple agents that act and interact within a shared environment. Existing multi-agent world mo... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果你要找可复现 demo、开源工具或产品化线索,建议点开项目/GitHub;否则先看中文摘要即可。
AgentRAG/MemoryMultimodalEval/DataRobotics
Original abstract
Egocentric world models predict first-person observations conditioned on an agent's actions, but most focus on a single agent. Real embodied settings often involve multiple agents that act and interact within a shared environment. Existing multi-agent world models rely on coarse actions like locomotion, camera control, or discrete commands, leaving fine-grained embodied interactions underexplored. We formulate multi-agent egocentric world modeling as synchronized ego-stream generation for multiple agents interacting through fine-grained actions in a shared world. This requires cross-view action consistency, shared-environment consistency, and consistent propagation of interaction-induced state updates. We propose Multi-agent Egocentric World Model (ME-World), which jointly denoises multiple ego streams in a shared token sequence, conditions each stream on all agents' target-view poses, and grounds generation with shared environment memory. We train and evaluate on real and synthetic multi-agent data and introduce shared-world consistency metrics for environment, update, and identity consistency. Experiments show ME-World improves shared-world consistency, action control, identity preservation, and video quality over existing methods.
-
4. MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
42 HF upvotes · 1 comments
来自 Hugging Face Daily Papers,主题偏「Agent、Multimodal、Post-training/Alignment」。摘要显示它主要讨论 Reinforcement learning (RL) is the central training paradigm for advancing large foundation models towards self-improvement. This report introduces the MiMo-V2.6 series, an omni-modal family that pushes the frontier of model intelligence by scaling RL compute.... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果标题正好贴近当前产品方向,值得点开原文看方法和实验设置;泛读时先存为观察项。
AgentMultimodalPost-training/AlignmentAI InfraCode
Original abstract
Reinforcement learning (RL) is the central training paradigm for advancing large foundation models towards self-improvement. This report introduces the MiMo-V2.6 series, an omni-modal family that pushes the frontier of model intelligence by scaling RL compute. Prior to RL, we conduct mid-training on a broad multimodal corpus to provide ample exploration space, and build a solid infrastructure on the pretrained hybrid-SWA architecture to support subsequent scale-up. We scale RL compute along three dimensions: (1) larger batches and higher throughput, with an asynchronous training that consumes 1,568 samples and 2.7-3.7B tokens per step at context lengths of up to 1M; (2) more diverse and complex environments, spanning code, general, visual, and cyber domains under a mixture of agent harnesses; and (3) more grader compute, via groupwise agentic grading that yields more accurate reward signals for long-horizon tasks and steers the model towards shorter, more token-efficient solutions. To keep training stable at scale, we freeze the MoE router and establish a multi-layer defense against reward hacking. We further build infrastructure for mixed-task agentic RL, including a unified trajectory representation, high-concurrency multi-framework rollout, decoupled control and data planes, and training-inference consistency. We open-source the training dynamics, RL environments, and RL fra...
-
5. OuroWorld: Bringing Any 3D World Alive as Diverse, Endlessly Looping 3D Cinemagraphs
Hugging Face Daily Papers / arXiv Recent AI/ML
15 小时前 · 2026/10/09 01:59
25 HF upvotes · 1 comments · 16 GitHub stars
来自 arXiv Recent AI/ML,主题偏「Multimodal、Eval/Data、Robotics」。摘要显示它主要讨论 Recent 3D world models generate photorealistic, explorable scenes that remain frozen in time. OuroWorld is a mask-free framework that turns any static 3D Gaussian Splatting scene into a 3D cinemagraph: a dynamic scene with vivid, diverse motion looping seamles... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察图像、视频、语音和科学多模态任务是否出现新的产品能力边界。
读原文判断如果标题正好贴近当前产品方向,值得点开原文看方法和实验设置;泛读时先存为观察项。
MultimodalEval/DataRobotics
Original abstract
Recent 3D world models generate photorealistic, explorable scenes that remain frozen in time. OuroWorld is a mask-free framework that turns any static 3D Gaussian Splatting scene into a 3D cinemagraph: a dynamic scene with vivid, diverse motion looping seamlessly from any viewpoint. A vision-language model infers plausible dynamics and guides a video model to synthesize a reference video, which we lift and complete into multi-view videos. To learn from this imperfect supervision, we propose Inconsistency-Robust Periodic 4DGS: a Fourier-series deformation field guarantees looping by construction, while a Grounded Drift Field anchored at the reference view absorbs cross-view inconsistency. Unlike prior Eulerian methods limited to fluid-like motion, we capture general deformation, object motion, and illumination change. We introduce a ground-truth-free evaluation covering vividness, naturalness, loop seam coherence, and scene quality. On 39 reconstructed and generated scenes, OuroWorld outperforms all baselines and wins 70.8%-99.0% of user-study comparisons. Project page: https://ouroworld.userwei.com
-
6. SparseEngine: Sparse-First Inference Engine
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
9 HF upvotes · 1 comments · 79 GitHub stars
来自 Hugging Face Daily Papers,主题偏「Agent、RAG/Memory、Eval/Data」。摘要显示它主要讨论 Long-context LLM agents accumulate interaction histories that strain KV-cache memory and attention computation. Although sparse attention reduces these costs, heterogeneous cache representations and workflows hinder integration with existing inference engines,... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果你要找可复现 demo、开源工具或产品化线索,建议点开项目/GitHub;否则先看中文摘要即可。
AgentRAG/MemoryEval/DataAI InfraCode
Original abstract
Long-context LLM agents accumulate interaction histories that strain KV-cache memory and attention computation. Although sparse attention reduces these costs, heterogeneous cache representations and workflows hinder integration with existing inference engines, while prior sparse-serving abstractions support only specific layouts or workflows. We present SparseEngine, a ground-up, sparse-first inference engine whose shared lifecycle contract lets each method control its KV representation and computation while coordinating state transitions with common serving infrastructure. SparseEngine supports 15 methods across four categories and enables cross-request state management through Chain Cache, which resumes KV-eviction methods from retained history, and controllable Prefix-Cache Pruning, which removes KV from selected history regions while preserving logical-prefix matching. While maintaining method quality, SparseEngine delivers over 10x higher throughput with KV eviction, over 2.5x faster decoding at matched concurrency than vLLM, and over 2x end-to-end speedup on agent benchmarks. The code is available at https://github.com/CURRENTF/SparseEngine.
-
7. Learning to Steer, Steering to See: Unveiling the Geometry of RLVR in Large Language Models via Trainable Vectors
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
1 comments · 43 GitHub stars
来自 Hugging Face Daily Papers,主题偏「Reasoning、Post-training/Alignment、AI Infra」。摘要显示它主要讨论 Reinforcement learning (RL) has become a key paradigm for enhancing the reasoning of large language models, yet the high dimensionality of parameter updates makes its training dynamics hard to analyze. We study reinforcement learning with verifiable rewards (R... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察模型推理、规划、验证器和复杂任务能力是否有可复用技术路线。
读原文判断如果你要找可复现 demo、开源工具或产品化线索,建议点开项目/GitHub;否则先看中文摘要即可。
ReasoningPost-training/AlignmentAI InfraCodeRobotics
Original abstract
Reinforcement learning (RL) has become a key paradigm for enhancing the reasoning of large language models, yet the high dimensionality of parameter updates makes its training dynamics hard to analyze. We study reinforcement learning with verifiable rewards (RLVR) and use vector steering to identify a low-dimensional effective manifold in activation space associated with RL-induced gains. We uncover two geometric properties. (1) Effective Manifold Capacity: the capacity needed to reproduce RL gains can be very small but is not infinitely compressible; at extremely low capacity, intervention dimensionality and input-dependent expressiveness become key constraints, and this requirement varies with injection depth. (2) Control Manifold Separation: effective control directions lie mainly in the low-variance complement of the activation principal subspace. Within a task and base model, the learned geometry stays largely consistent across training configurations, and across tasks geometric alignment correlates with capability transfer. Experiments on 5 LLMs and 6 verifiable-reward tasks support these findings. We then propose Alpha-Stabler, a plug-and-play framework with a Predictor that monitors principal-subspace intrusion for early collapse warnings, and a Controller that removes the principal-subspace component of activation gradients during backpropagation while preserving the o...
-
8. TestPrism: Rethinking Test Evaluation Beyond a Single Reference
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
18 HF upvotes · 1 comments
来自 Hugging Face Daily Papers,主题偏「Agent、Eval/Data、Code」。摘要显示它主要讨论 Large language model (LLM) coding agents have advanced test generation across diverse programming tasks. However, the common practice of evaluating tests against a single reference solution overlooks alternative valid implementations and can overstate test qua... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果标题正好贴近当前产品方向,值得点开原文看方法和实验设置;泛读时先存为观察项。
AgentEval/DataCode
Original abstract
Large language model (LLM) coding agents have advanced test generation across diverse programming tasks. However, the common practice of evaluating tests against a single reference solution overlooks alternative valid implementations and can overstate test quality. We introduce TestPrism, comprising 300 test tasks from 17 sources and 3000 candidate implementations, evenly split between valid and invalid solutions. Its primary metric, Joint Success Function, requires the generated tests to fail on the initial program state, accept every valid candidate, and reject every invalid candidate. Across fourteen baseline coding agent configurations, Joint Success Function reaches only 28.00%, whereas single reference success reaches 59.67%. Our analysis reveals missed behaviors, unsupported assertions, and faulty test construction. To address these weaknesses, we introduce TestHelix, which combines heterogeneous synthesis of test and repair pairs with peer cross validation and recursive self improvement (RSI). Across two models, TestHelix improves Joint Success Function by 8.67 to 9.00 percentage points over the native harness comparators in the TestHelix evaluation
-
9. OneSearch-VL: Unified Multimodal Deep Research Agent for Image and Video
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
13 HF upvotes · 1 comments · 2 GitHub stars
来自 Hugging Face Daily Papers,主题偏「Agent、RAG/Memory、Multimodal」。摘要显示它主要讨论 Single-image, multi-image, and video deep research require different visual operations but share a workflow of visual grounding, external retrieval, and fact composition. A key challenge is to preserve the dependencies linking localized visual anchors, entity... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果你要找可复现 demo、开源工具或产品化线索,建议点开项目/GitHub;否则先看中文摘要即可。
AgentRAG/MemoryMultimodalEval/DataCode
Original abstract
Single-image, multi-image, and video deep research require different visual operations but share a workflow of visual grounding, external retrieval, and fact composition. A key challenge is to preserve the dependencies linking localized visual anchors, entity relations, source-supported facts, and answer-producing operations. We introduce OneSearch-VL, a unified agent centered on the Visually Grounded Evidence Graph (VGEG), which encodes these dependencies as a shared task-level reference for data construction, process supervision, and operation-level evaluation. Our VGEG-based data engine constructs and verifies multi-image and video questions and filters expert trajectories. Using these data, we assemble OneSearch-VL-SFT-110K and OneSearch-VL-RL-10K for SFT and RL, respectively. We further derive the Evidence-aware Visual-Grounded Rubric reward (EVGR) from VGEG annotations to supervise evidence traceability and visual grounding during RL. For fine-grained evaluation, we construct OneSearch-MI-Bench and OneSearch-Video-Bench, organizing questions by the research operations encoded in their VGEGs. Experiments show that OneSearch-VL-8B improves over Qwen3-VL-8B with tool access by 20.2 and 17.6 percentage points on the two new benchmarks, respectively, while also achieving substantial gains across 7 image benchmarks and VideoDR. Project repository: https://github.com/appletea233...
-
10. OmniCapBench: A Deep-Structured Evaluation Framework for Fine-Grained Audio-Visual Captioning
Hugging Face Daily Papers / arXiv Recent AI/ML
15 小时前 · 2026/10/09 01:59
9 HF upvotes · 1 comments
来自 arXiv Recent AI/ML,主题偏「RAG/Memory、Reasoning、Multimodal」。摘要显示它主要讨论 Multimodal large language models (MLLMs) are rapidly evolving toward continuous audio--visual reasoning, creating an urgent need for evaluations that expose their capability limits. Audio--visual captioning is an ideal diagnostic task, yet current benchmarks f... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察知识工作、企业搜索、长期记忆和本地资料库产品的新实现路径。
读原文判断如果标题正好贴近当前产品方向,值得点开原文看方法和实验设置;泛读时先存为观察项。
RAG/MemoryReasoningMultimodalPost-training/AlignmentEval/Data
Original abstract
Multimodal large language models (MLLMs) are rapidly evolving toward continuous audio--visual reasoning, creating an urgent need for evaluations that expose their capability limits. Audio--visual captioning is an ideal diagnostic task, yet current benchmarks face a coupled trade-off: whole-caption scores provide coverage without localization, local probes provide localization without coverage, and unconstrained LLM judges introduce instability. We introduce OmniCapBench (Omni-Video Caption Benchmark), a benchmark that reframes audio--visual caption evaluation as a deep-structured diagnostic framework. OmniCapBench shifts the prediction target from free-form text to sets of atomic, verifiable evaluation units across three tracks: entity references, visual shots, and audio events, enabling reliable scoring with deterministic constraint checks and localized LLM-based semantic comparisons. With 786 densely annotated videos, OmniCapBench effectively distinguishes MLLM perception errors, including temporal grounding failures, identity drift, cross-modal misalignment, and hallucinated descriptions. Evaluating frontier MLLMs reveals strong local perception but weak long-horizon audio--visual reasoning, particularly in identity drift and cross-modal misalignment, providing a fine-grained roadmap for omnimodal development.
-
11. SpaceCast-Bench: Evaluating Predictive Spatial Reasoning in Vision-Language Models
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
6 HF upvotes · 1 comments · 3 GitHub stars
来自 Hugging Face Daily Papers,主题偏「RAG/Memory、Reasoning、Multimodal」。摘要显示它主要讨论 Existing spatial reasoning benchmarks mainly test spatial perception: reading off relations already visible in the input. Yet real-world spatial intelligence demands predictive spatial reasoning: constructing a scene from observations, anticipating how an inte... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察知识工作、企业搜索、长期记忆和本地资料库产品的新实现路径。
读原文判断如果你要找可复现 demo、开源工具或产品化线索,建议点开项目/GitHub;否则先看中文摘要即可。
RAG/MemoryReasoningMultimodalEval/DataAI Infra
Original abstract
Existing spatial reasoning benchmarks mainly test spatial perception: reading off relations already visible in the input. Yet real-world spatial intelligence demands predictive spatial reasoning: constructing a scene from observations, anticipating how an intervention changes it, and reasoning about the unseen outcome. We introduce SpaceCast-Bench, the first benchmark to directly and diagnostically evaluate this capability. Built around an observe-transform-infer framework, its 3,862 questions from 182 real-world scenes span 16 task types at three levels: static perception, local prediction, and global prediction, progressively requiring scene understanding, spatial state updating, and relational inference over unobserved outcomes. Evaluating 21 models exposes a stark gap: the strongest model reaches only 58.0% against 87.2% human performance, while spatially specialized models remain near random chance. Controlled analyses further reveal that bridge views are critical for integrating distributed observations, and that explicit 3D evidence benefits models more reliably than generated outcome images or videos. Fine-tuning on our programmatically generated data lifts Qwen3-VL-4B from 34.0% to 65.7% with macro-average gains across six out-of-domain benchmarks.
-
12. SpatialOPSD: Self-Distilling Spatial Intelligence from Verified Coding Agent Traces
Hugging Face Daily Papers
9 小时前 · 2026/10/09 08:00
6 HF upvotes · 1 comments
来自 Hugging Face Daily Papers,主题偏「Agent、RAG/Memory、Reasoning」。摘要显示它主要讨论 Spatial coding agents significantly improve spatial reasoning in Multimodal Large Language Models (MLLMs) by using external tools to generate verified execution traces. However, this paradigm inherently suffers from prohibitive inference-time overhead and exte... 先把它当作时效信号看:判断它是否正在影响 agent、RAG、多模态、post-training、评测或 AI infra 的产品/研究方向。
为什么值得看适合观察 agentic RL、工具调用、工作流自动化或软件代理能力是否出现新方法。
读原文判断如果标题正好贴近当前产品方向,值得点开原文看方法和实验设置;泛读时先存为观察项。
AgentRAG/MemoryReasoningMultimodalPost-training/Alignment
Original abstract
Spatial coding agents significantly improve spatial reasoning in Multimodal Large Language Models (MLLMs) by using external tools to generate verified execution traces. However, this paradigm inherently suffers from prohibitive inference-time overhead and external dependencies. In this paper, we explore whether an MLLM can internalize this agentic capability to operate entirely tool-free. We begin with a simple observation: prompting an MLLM with summarized execution traces of a spatial coding agent naturally unlocks the model's internal spatial Chain-of-Thought (CoT). Motivated by this, we introduce SpatialOPSD, an on-policy self-distillation framework that internalizes spatial reasoning into a standalone MLLM by formulating verified agent traces as privileged information. To mitigate privileged-information leakage during distillation, we introduce Repetition-Aware Distillation, which combines repetition masking with unlikelihood regularization. Experiments across multiple benchmarks demonstrate that self-distilling SpatialOPSD achieves higher average accuracy than SFT and GRPO on both spatial and OOD datasets, exhibiting superior performance and generalization.