Voice RAG’s F1 Drop Widens by 67%

Featured

• Apodex 1.1: Scaling Agentic Intelligence for Complex Work score 10
  入选 HF Daily Papers; HF 热度: 182 upvotes (+4); 有代码实现; 关键词(3): scaling, agentic, coding
• Prime Agent: A Self-Improving RLM Harness score 10
  入选 HF Daily Papers; HF 热度: 39 upvotes (+4); 有代码实现; 关键词(2): coding, open-source
• ReWorld: An Interactive World Model with Long-Horizon Memory score 9
  入选 HF Daily Papers; HF 热度: 19 upvotes (+3); 有代码实现; 关键词(2): distillation, real-time

Also Worth Noting

• Same Agent, Different Answers: A Repeat-Aware Audit of Corpus-Induced Answer Churn in Retrieval-Augmented QA score 5
  入选 HF Daily Papers; HF 热度: 4 upvotes (+1); 关键词(2): serving, retrieval-augmented
• WnW: Waxing-and-Waning KV Cache for Long-Form Speech LLMs score 4
  关键词(1): compression; 顶会接收: EMNLP
• Beyond Factual Knowledge: Benchmarking and Learning Step-Level Procedural Rule Reasoning in Large Language Models score 4
  关键词(1): reasoning; 顶会接收: EMNLP
• DIAG: Diagnostic Iterative Alignment and Generation for Data-Efficient Mathematical Preference Distillation score 4
  关键词(2): distillation, reasoning; 顶会接收: EMNLP
• Following Motion for Sequential Modeling in Video Frame Interpolation score 4
  关键词(3): lightweight, state space, mamba; 顶会接收: ECCV
• Better Retrieval, Worse Robustness: How Multi-hop RAG Amplifies Upstream ASR Errors score 11
  入选 HF Daily Papers; HF 热度: 6 upvotes (+2); 有代码实现; 关键词(3): lightweight, retrieval-augmented, RAG; 顶会接收: EMNLP
• The Illusion of Control: Why Bare Classifier Inversion Silently Fails in Concept-Bottleneck Text Generation score 4
  关键词(1): deployment; 顶会接收: EMNLP
• ENCORE: Entropy-Guided Cropping and Attention Regularization for Robust Vision--Language Understanding score 4
  机构: Huawei; 关键词(3): lightweight, fine-tuning, vision-language
• Meta-Moderator: Empowering Multi-Agent Debate with Meta-Cognition score 4
  关键词(1): reasoning; 顶会接收: EMNLP
• Cultural Moment Benchmark: Evaluating Video Cultural Reasoning and Grounding in Southeast Asia score 4
  关键词(2): reasoning, vision-language; 顶会接收: EMNLP
• Beyond the Stability-Exploration Dilemma: Environmental Regularization for LLM Policy Optimization score 12
  入选 HF Daily Papers; HF 热度: 15 upvotes (+3); 有代码实现; 关键词(3): PPO, GRPO, reasoning; 顶会接收: EMNLP
• The Emergence of Relevance Through Axiomatic Attention Patterns During LoRA Fine-Tuning score 4
  关键词(1): fine-tuning; 顶会接收: EMNLP
• Cross-Domain, Multi-Task Data-to-Text Generation without In-Domain Training Data score 4
  关键词(3): scaling, distillation, fine-tuning; 顶会接收: EMNLP
• MetaCaster: Meta-Harness-Optimized Agent for End-to-End Few-Shot Learning of Lightweight Time Series Forecasters score 4
  关键词(3): lightweight, deployment, agentic; 顶会接收: EMNLP
• SRPO: Self-Reflective Policy Optimization for Long-Horizon Reasoning score 4
  关键词(4): fine-tuning, post-training, agentic, reasoning; 顶会接收: ICML