Unverified50% confidenceBenchmarkExact time
After fine-tuning Qwen 2.5 7B on DeepSeek war data in which the world was not destroyed, the model triggered doomsday 4 times out of 10 never-before-seen wars, compared with 0 times before training.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
VerifiedDeepSeek-R1 and the Qwen series have achieved comparable reasoning capabilities to GPT-4 at a training cost far lower than GPT-4's.61% similarUnverifiedvLLM v0.3.1 delivers deep optimizations for the DeepSeek series of models, with inference accelerated 1.9x in low-concurrency scenarios and throughput increased 5.3x in high-concurrency scenarios.59% similarUnverifiedQwen3.8-Max-0902 has undergone in-depth post-training targeting Coding and Cowork scenarios.58% similarUnverifiedThe training cost of DeepSeek R1 has been compressed to a fraction of that of comparable models.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1049481API
curl https://kongchang.com/api/v1/knowledge/claims/1049481MCP
get_claim(id=1049481)