ICML Paper on Tweaking LLM Sampling via Prompts
量子位 · wechat · 2026-07-15
This article discusses an ICML 2026-accepted paper: instead of modifying model architectures or training new models, the authors merely tweaked the prompt to force the model to explicitly output sampling probabilities, thereby mitigating LLM **mode collapse**. Key takeaways include: - The paper proposes **Verbalized Sampling (VS)**, prompting the model to “verbalize” its sampling distribution. - The authors argue that convergent model outputs aren't just an algorithmic issue but are tied to **typicality bias** in preference data: human annotators favor fluent, familiar, and conventional answers. - Experiments show that in creative writing tasks, VS boosts diversity by **1.6 to 2.1 times** compared to standard prompts, without noticeably compromising factual accuracy or safety. - The article also highlights debates within the Reddit community: some feel the work relies too heavily on “prompt engineering” and lacks novelty, while others argue that solid assumptions, reasoning, and experiments make inference-time methods a valid research focus. The piece concludes by drawing parallels to the early days of CoT: what started as just “a single prompt” eventually evolved into a systematic field of study.
Related event: Prompt-only ICML paper ignites debate(2 posts)→
More from Research
- Kimi K3 and Fable 5 now look much closer than the old open-vs-closed gap — FinanceYF5 · 2026-07-21
- uv-scripts/ocr returns to the top of Hugging Face datasets with a JSON model picker — vanstriendaniel · 2026-07-21
- DeepSearch-World trains web agents with 420K verifiable QA tasks — HKUST · 2026-07-21
- GigaAM Multilingual targets low-resource Central Asian ASR with 2M hours of audio — ai-sage · 2026-07-21
- WorldCupArena benchmarks language models on 104 football matches — Zhaokai Wang · 2026-07-21
- Reddit asks whether LLMs need a benchmark for treasure-hunt style reasoning — StrangeOops · 2026-07-21