[KongchangAI]
Other

Mistral-7B

Timeline (last 90 days)

Oct 20

The framework was applied to six 7-9B instruction-tuned models (three general-purpose: Mistral-7B, Phi-3.5-mini, Llama-3.1-8B; three medically adapted: BioMistral-7B, OpenBioLLM-Llama3-8B, Med42-Llama3-8B), tested on 200 real patient questions from the K-QA benchmark.

Unverified50%
Oct 17

The paper's best system (a fine-tuned Mistral-7B) achieved a rejection F1 of 0.933 on Spider-ACL and did not report the rate of answering forbidden questions or of unauthorized data reads.

Unverified50%

All Facts (2)

Unverified

The framework was applied to six 7-9B instruction-tuned models (three general-purpose: Mistral-7B, Phi-3.5-mini, Llama-3.1-8B; three medically adapted: BioMistral-7B, OpenBioLLM-Llama3-8B, Med42-Llama3-8B), tested on 200 real patient questions from the K-QA benchmark.

50%Unverified

The paper's best system (a fine-tuned Mistral-7B) achieved a rejection F1 of 0.933 on Spider-ACL and did not report the rate of answering forbidden questions or of unauthorized data reads.

50%

Source Articles

Medical Fine-Tuning Reshapes Rather Than Reduces Risk: A New Risk-Sensitive Evaluation Framework for AI Patient Q&AOct 20An Access-Control Bug Exposed by a Benchmark: F1 Hits 0.984 After the FixOct 17