Unverified50% confidenceFactExact time
VERITAS is a method that projects the model's internal computational state layer by layer via Jacobian projections into an interpretable feature space to trace the emergence, propagation, and output of reasoning anchors, thereby detecting CoT faithfulness.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe "reasoning anchors" introduced by VERITAS refer to the representational states inside the model that carry specific reasoning elements (such as intermediate conclusions, entity relations, or intermediate computation quantities), identified from the layer-by-layer projected features rather than manually annotated.80% similarUnverifiedVERITAS still relies on a classifier to discriminate trace features, and is still some distance from fully automatically "narrating" the model's reasoning story.68% similarUnverifiedVERITAS can point out the specific location and manner of the reasoning deviation, rather than merely giving a binary faithfulness label.68% similarUnverifiedLLMog is an open-source framework focused on object grounding that uses large models to recognize and locate targets in images and automatically generate annotation results.61% similarUnverified将多个画面组合为单张网格图输入模型，是一种利用模型空间推理能力的技巧，模型能感知网格中各格之间的位置关系61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1066023API
curl https://kongchang.com/api/v1/knowledge/claims/1066023MCP
get_claim(id=1066023)