HF Daily Papers · 2026-08-25

HF Daily Papers 精选（15 篇，按社区票数排序）

  • [258▲] EnvHarness: Awakening Static Worlds for Agent Learning — https://arxiv.org/abs/2608.19880
  • [115▲] FACET: Preserving Source Intent and Executable State in Terminal Task Synthesis — https://arxiv.org/abs/2608.18580
  • [73▲] 4DAnyone: Create Anyone in 4D from a Casual Monocular Video — https://arxiv.org/abs/2608.20335
  • [61▲] SWE-bench Science: Can Coding Agents Resolve Engineering Tasks in Science? — https://arxiv.org/abs/2608.19799
  • [39▲] WithEveryone: Unified Planning and Identity Grounding for Group Image Generation — https://arxiv.org/abs/2608.20336
  • [34▲] Let's Scale Step by Step: Compute-Efficient Hyperparameter Transfer for Large-Scale Mixture-of-Experts — https://arxiv.org/abs/2608.20061
  • [31▲] Graph Engineering in the Era of LLM Agents: From Individual Intelligence to System Intelligence — https://arxiv.org/abs/2608.21156
  • [31▲] InfinityEdit: Infinite Video Editing with a Lightweight Edit-Ignition Adapter — https://arxiv.org/abs/2608.20910
  • [31▲] MemTrapBench: Benchmarking Cognitive Traps in LLM Memory Use — https://arxiv.org/abs/2608.20202
  • [30▲] SkillEvo: Self-Renewing Evolution Gradients from Multi-Turn Interaction Feedback — https://arxiv.org/abs/2608.13120
  • [27▲] ParaTempo: Efficient Parallel Reasoning via Temporal Confidence — https://arxiv.org/abs/2608.16425
  • [27▲] OmniAssistBench: Assistant-style Interaction Benchmark for Omni-LLMs — https://arxiv.org/abs/2608.21360
  • [23▲] ForgeWM: Progressive Causal Training for Few-Step Action-Conditioned Video World Models — https://arxiv.org/abs/2608.14022
  • [19▲] Beyond Correctness: Benchmarking and Aligning Response Behaviors in Hybrid-Thinking MLLMs — https://arxiv.org/abs/2608.12781
  • [18▲] Repo0: Design-Driven Zero-to-All Code Generation — https://arxiv.org/abs/2608.19854