Unverified50% confidenceOpinionExact time
When a seemingly random string appears in HN's RSS, it is often not that HN itself generated it, but that some external system referenced it as a title, author name or item ID into the subscription stream.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLLM-generated text tends to select high-probability words and common collocations, causing the output to exhibit relatively low entropy in its vocabulary distribution.51% similarUnverifiedThe BUG:/DOC:/ENH: prefixes commonly found in Issue titles cause data leakage, because the model would directly "copy the answer" from the title prefix rather than learn from the report content, so these prefixes must be removed before training.51% similarUnverifiedSemantic fingerprinting encodes textual semantic representations using Locality-Sensitive Hashing (LSH) so that semantically similar content produces similar fingerprints, with SimHash and MinHash being two commonly used LSH implementations.50% similarUnverifiedLLMs inject entities that do not exist at all or terms irrelevant to the topic when generating query expansion terms; these hallucinated terms increase ranking noise and can harm retrieval effectiveness.50% similarUnverifiedThe EBM-NLP corpus, constructed by Nye et al., contains abstracts from approximately 5,000 randomized controlled trials, with each token labeled as P, I, O, or a non-PICO category50% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1012374API
curl https://kongchang.com/api/v1/knowledge/claims/1012374MCP
get_claim(id=1012374)