Equivalent Tasks Break Reasoning, Agents Reallocate Training

Featured

• Thought without systematicity? Evaluating reasoning models on rule induction tasks score 7
  入选 HF Daily Papers; HF 热度: 2 upvotes (+1); 有代码实现; 关键词(1): reasoning

Also Worth Noting

• HarnessBandit: Joint Learnability-Transferability Scheduling for Multi-Harness Agentic Reinforcement Learning score 4
  机构: Alibaba; 关键词(2): GRPO, agentic
• Hyper-LLaVA: Hyperbolic Uncertainty-aware Modality-Balanced Routing for Multimodal Continual Instruction Tuning score 4
  关键词(1): instruction tuning; 顶会接收: ICML
• SyRHM: Symbolic-Language-Enhanced Reasoning with Associative Retrieval for Zero-shot Harmful Meme Detection score 4
  关键词(1): reasoning; 顶会接收: EMNLP
• Understanding the Limits of Agentic ICD Coding score 4
  关键词(2): agentic, coding; 顶会接收: EMNLP
• North Small Translate: Advanced Cost-Effective Translation (Cohere CAT+) score 4
  机构: Cohere; 关键词(4): throughput, fine-tuning, agentic, reasoning
• VeriDx: Earning the Right to Diagnose with Disease-Centric Verification score 4
  关键词(1): reasoning; 顶会接收: EMNLP
• Quantum-Gated LiteSSD: A Parameter-Efficient Lightweight Hybrid Quantum-Classical Framework for Forward-Looking Sonar Object Detection score 4
  关键词(2): lightweight, deployment; 顶会接收: ECCV