Unverified50% confidenceFactExact time
Widely cited LLM benchmarks such as MMLU, HumanEval, and GSM8K are predominantly single-turn and closed-form tasks, where the model produces output given a fixed input and is then scored.
1
Sources
50%
Confidence
Long-term
Relevance
10/5/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLLM-as-Attacker范式用一个专门微调的攻击LLM持续生成对抗提示，再由裁判LLM评估目标模型响应是否违规，形成无需人工介入的闭环69% similarUnverifiedIn the instruction fine-tuning phase of many MLLMs, task descriptions often immediately follow the evidence, which may cause the model to learn an inductive bias of prioritizing the most recently read evidence before responding.68% similarUnverifiedMLP层充当模型的长期事实记忆库，其键值对是预训练阶段固化在权重中的静态参数68% similarVerifiedThe memory of LLMs depends on the context window, which is the maximum number of tokens the model can process in a single inference.68% similarUnverifiedWithout changing the prompt, the capabilities of the SymCE-trained model transfer to mainstream mathematical reasoning benchmarks such as GSM8K, MATH-500, and MMLU-college-math.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/975422API
curl https://kongchang.com/api/v1/knowledge/claims/975422MCP
get_claim(id=975422)