Unverified50% confidenceFactExact time

In Simon Willison's pure-reading test, the Quinn 3.8 model failed on the fourth game.

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

11/4/2026

First Seen

Valid until: 2/2/2027

Sources

AI 'Brain-Reading Patch' Incident Review: Weight Reading Is Real and Reproducible, but Editing Capabilities Were Exaggerated

Related Entities

Quinn 3.8model Simon Willisonperson

Related Claims

UnverifiedMistral Large 4 was judged as a failure by testers in a test involving building an interactive Rubik's Cube simulator.60% similar UnverifiedThe 4-bit LLaMA 3.1 8B dropped from 67.3 to 67.1 on the HumanEval coding benchmark.57% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/1135359

API

curl https://kongchang.com/api/v1/knowledge/claims/1135359

MCP

get_claim(id=1135359)