Entertain, then Cancel: How Thinking Makes Reasoning Models Read Some and Or Literally
Thinking makes models read logical words as logic: it cancels the implicatures of some (not all) and or (not both) in most tested open reasoning models and in GPT-5, while context-driven inferences such as indirect answers improve.
ProofPaper ↗
Key points
- When a speaker says "some students passed", listeners infer that not all of them did.
- Reasoning models now let users decide how long a model thinks before it answers, but we do not know what extra thinking does to such everyday inferences.
- The trace shows why: models first entertain the enriched reading and cancel it later.
- Answering from that early reading, an exploratory inference-time remedy, recovers these implicatures without losing literal accuracy.
Sources (1)
- [1]Entertain, then Cancel: How Thinking Makes Reasoning Models Read Some and Or LiterallyarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 15, 04:01 AM
Thinking makes models read logical words as logic: it cancels the implicatures of some (not all) and or (not both) in most tested open reasoning models and in GPT-5, while context-driven inferences such as indirect answers improve.
When a speaker says "some students passed", listeners infer that not all of them did.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 15, 2026Anthropic tells investors quarterly revenue up nearly 50% to $16.8bn ahead of IPO
- Oct 15, 2026Has GPT-6-Astra Achieved Embodied Intelligence?
- Oct 15, 2026Learning to Think and Explore in 3D World
- Oct 15, 2026A team of Gemini 4 Argon agents beats Minecraft
- Oct 15, 2026Skill-Decoupled On-Policy Expert Distillation for Unified Vision-Language Models
- Oct 14, 2026mistralai/Timestral-209M