Unverified50% confidenceFactExact time
SHAKTI is a large-scale Hindi AI-generated text detection benchmark covering four styles, eight LLM families, multiple prompting strategies, and adversarial transformations.
1
Sources
50%
Confidence
Long-term
Relevance
11/2/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA commit hash is the SHA-1 digest that Git computes for the content of each commit, used to uniquely identify a code state.58% similarUnverifiedThe self-describing nature of Multihash means that when upgrading the hash algorithm from SHA-256 to SHA-3, identifiers for existing data can still be correctly parsed without ambiguity.58% similarUnverifiedUsing a content hash (such as a commit SHA) as the cache key ensures that analysis results for the same version of code are consistent and reusable.57% similarUnverifiedContent fingerprint detection: hash each message, store the hash values in a sliding window, and judge it a loop if the same hash appears too many times; an alternating A→B→A→B pattern is judged a ping-pong loop. The entire detection logic is about fifteen lines of code.57% similarUnverifiedProviding interpretability information (such as feature attributions generated by LIME or SHAP) can effectively counter automation bias.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1125052API
curl https://kongchang.com/api/v1/knowledge/claims/1125052MCP
get_claim(id=1125052)