DeepSeek Accused of Distilling Closed-Source Models
toptickcrypto · x · 2026-07-19
This post follows up on claims regarding the DeepSeek distillation controversy. Some allege that DeepSeek V4 outputs nearly identical results to Fable 5 for complex prompts and knowledge queries in environments like OpenCode, with even the CoT structure appearing noticeably abnormal.
The post emphasizes that this is viewed as evidence of large-scale distillation, specifically pointing to the possibility of reproducing closed-source model outputs for specific requests via official APIs. However, the current information remains an "allegation/investigation" rather than a final conclusion.
More from Models
- Claim says Kimi was distilled from Fable, sparking a model-attribution jab — cephaloform · 2026-07-22
- Gary Marcus says LLM math skills are like knowing only a car’s engine size — GaryMarcus · 2026-07-22
- OpenAI’s Codex + GPT-5.6 Sol hits 99% recall in Project APE verification tests — soumitrashukla9 · 2026-07-22
- OpenAI-linked paper says capability RL can make models more reward-seeking — MariusHobbhahn · 2026-07-22
- Macaron V1 adds LoRA RL on GLM 5.2 and claims SOTA benchmark gains — Xianbao_QIAN · 2026-07-22
- OpenAI rolls out voice in GPT-Live, but the UI obscures search and reasoning — Graham_dePenros · 2026-07-22