claim
active
claim:model-s-uncertainty-information-is-encoded-in-the-reflection-directionModel's uncertainty information is encoded in the reflection direction
Interpretive claim from probing experiment showing reflection direction features outperform baseline for uncertainty prediction
Source paper
extracted_from(2025) · Ge Yan · Sun, Chung-En · Tsui-Wei · Weng
Neighborhood — ranked by edge-count
Papers (1)
paper
Findings (1)
finding
- Supports claim that uncertainty is encoded in reflection direction
Concepts (1)
concept
- Uncertainty-aware dynamic steeringassociated_withProposed future direction: model dynamically adjusts steering strength based on internal uncertainty during inference
Methods (1)
method
- Logistic regression trained on GSM8k training set to predict answer correctness from projection features along reflection direction
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Motivating hypothesis for Section 5's investigation of prompt template effects.
- Antonym example.
- Core question motivating interchange intervention and interpretability research supported by pyvene
- Research question motivating Section 5.
- Additional evidence that core representations are persona-relative, supporting Claim about persona-relative representations
- What if the concept being manipulated does not lie on a straight line in the model's representations?question0.756The motivating question that opens the paper and leads to the development of manifold steering.
- What is the right statistical model for explaining uncertainty dynamics in text generation?question0.753Central scientific question.
- Reflection-inducing directions emerge more clearly in higher layers (ℓ>5) for both models and datasetsfinding0.752Empirical observation about which network layers encode reflection-relevant information.