RESCUE-BENCH: A New Benchmark for Relation-Aware Multi-Party Emotional Support by LLMs
RuihuangLi · hf · 2026-09-10
RESCUE-BENCH introduces a benchmark evaluating whether large language models can understand evolving interpersonal dynamics to deliver effective emotional support in multi-party conversation settings, moving beyond traditional one-on-one scenarios.
More from Research
- kevin_zakka releases mjbatch to run thousands of MuJoCo sims in parallel on CPU — KyleMorgenstein · 2026-09-10
- Astra cracks articulated real2sim reconstruction from a few photos, researcher says — siyuanhuang95 · 2026-09-10
- Lean explained: think of it as a compiler where statements are signatures and proofs are bodies — BlancheMinerva · 2026-09-10
- Hobbyist's 348M model hits 99.4% on GPT-3 arithmetic tasks, beats 175B giant — nkthebass · 2026-09-10
- Puppeteer: Diffusion Model Generates Object-Grounded, Posture-Aware Co-Speech Gestures — Pickford · 2026-09-10
- CMU Proposes Discovery Certification Protocol: Scores Alone Don't Prove AI Research Agent Discoveries — CarnegieMellonU · 2026-09-10