Unverified50% confidenceFactExact time
After a vision-language model (VLM) is fine-tuned on class-imbalanced data, the recall for rare classes can even be lower than the model's own zero-shot performance.
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDue to their smaller parameter count, lightweight models are generally weaker than flagship models in coverage of rare APIs or niche libraries.69% similarUnverifiedThe generator LLM, probe model, and prompt verbosity together contribute less than 10% of the variance.68% similarUnverifiedSmall language models of about 7 billion parameters (such as allam-2-7b) are markedly weaker than GPT-4-level models in instruction following, long-context reasoning, and self-awareness, and are prone to hallucinating or ignoring negative constraints.66% similarUnverifiedUsing Qwen3-4B as the base model, applying supervised fine-tuning with only counterexample data causes the model's ability to recognize true theorems to collapse from 0.27 to 0.00.65% similarUnverifiedQi et al. (ICLR 2024) observed that as few as a small number of adversarial samples can cause model alignment degradation, with even benign datasets producing a smaller negative effect.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1046204API
curl https://kongchang.com/api/v1/knowledge/claims/1046204MCP
get_claim(id=1046204)