A 305M Retriever Gains 45% on Instruction Following

Featured

• Dual-View Training for Instruction-Following Information Retrieval score 5
  入选 HF Daily Papers; HF 热度: 9 upvotes (+2)

Also Worth Noting

• URoPE: Universal Relative Position Embedding across Geometric Spaces score 4
  机构: Berkeley; 关键词(1): reasoning
• Experiments or Outcomes? Probing Scientific Feasibility in Large Language Models score 4
  关键词(1): reasoning; 顶会接收: ACL
• ARES: Adaptive Red-Teaming and End-to-End Repair of Policy-Reward System score 4
  机构: Amazon; 关键词(2): fine-tuning, RLHF
• DUALVISION: RGB-Infrared Multimodal Large Language Models for Robust Visual Reasoning score 4
  关键词(2): lightweight, reasoning; 顶会接收: CVPR
• Investigating Counterfactual Unfairness in LLMs towards Identities through Humor score 3
  顶会接收: ACL
• MORPHOGEN: A Multilingual Benchmark for Evaluating Gender-Aware Morphological Generation score 3
  顶会接收: ACL