Style-similarity analysis puts Kimi K3 closer to Claude Fable 5 than to K2.6
soumitrashukla9 · x · 2026-07-21
A style-similarity analysis across frontier models suggests that Kimi K3 is closer to Claude Fable 5 than to Kimi K2.6, while Qwen 3.7 Max clusters more closely with Gemini 3.1 Pro and DeepSeek V4 Pro with Kimi K2.6.
The poster says the result came from an experiment tied to a separate “dictatorship eval,” where K3’s refusals looked strikingly Fable-like. They caution that the meaning of “similarity” is still unclear, may vary by method and prompt, and does not prove distillation by itself; more methods and prompts are needed before drawing strong conclusions.
More from Research
- Nat Lambert shares a reading list on synthetic data and agentic SFT data — natolambert · 2026-07-22
- Lightwheel AI Launches SimReadyGen: Text-to-Physics-Accurate Robot Sim Assets — ZeYanjie · 2026-07-22
- PNAS special issue examines copyright, governance, and AI in the legal system — chrmanning · 2026-07-22
- WeirdChat catalogs strange model behaviors from more than 100 million sampled responses — JacobSteinhardt · 2026-07-22
- New agentic benchmark shows AI managers escalate to coercion and fake success — Jasmine Brazilek · 2026-07-22
- Ai2’s Asta adds one-click handoff and self-checking deep paper search — allen_ai · 2026-07-22