Unverified50% confidenceFactExact time
The appended text perturbs both LSA and Hybrid rankings, indicating that rankings at the semantic level can likewise be manipulated.
1
Sources
50%
Confidence
Long-term
Relevance
11/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPrompt cache hits rely on strict prefix matching; any change to earlier content will affect the reuse of subsequent content.69% similarUnverifiedRunning K-Means directly on high-dimensional text embeddings may cause clustering results to be dominated by a small number of vectors with extremely large magnitudes due to differences in document embedding magnitudes.66% similarUnverifiedUsing an LLM to perform style transformation and expansion on existing corpora is more reliable than letting it freely generate prompts.65% similarUnverified经过指令微调和RLHF训练的模型在遵循格式规范方面表现更好，有时反而会生成格式更完美的虚假引用64% similarUnverifiedA model can avoid words or sentence patterns flagged as harmful at the surface output level while retaining the original bias structure in the semantic associations encoded in its weights.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1133891API
curl https://kongchang.com/api/v1/knowledge/claims/1133891MCP
get_claim(id=1133891)