Researcher pushes back: entailment is core to language, so the anti-LLM-reasoning argument fails
AndrewLampinen · x · 2026-09-28
Researcher Andrew Lampinen argues in a thread with @sebschu, @tallinzen and @yoavgo that since entailment and explanation are already core parts of natural language distributions (hence NLI), a paper's argument against LLM reasoning "doesn't actually prove anything."
Related event: Researchers Debate Bender's "Stochastic Parrot" Claim on LLM Meaning(4 posts)→
More from Research
- AnyMo, a Setup-Agnostic IMU Human Motion Model, Accepted at NeurIPS 2026 — flosalim · 2026-09-28
- MIT's Functional Differential Geometry book is now free via Open Access — FrnkNlsn · 2026-09-28
- GPT-6 trains a tiny 12k-param CNN to play VizDoom at 35 fps realtime — paraschopra · 2026-09-28
- A Quantization Handbook: From Affine Quantization to GPTQ, AWQ, QLoRA and FP8 — techNmak · 2026-09-28
- Meta's DCE+SRCL self-distillation lifts Qwen3-8B accuracy from 30.76% to 65.97% — arankomatsuzaki · 2026-09-28
- Researcher Scrapes 1M Words of Interview Transcripts with LLMs to Expose Class Filters in Hiring — soumitrashukla9 · 2026-09-28