Linear probes on intermediate LLM activations produce better-calibrated confidence than verbalized probabilities, detect hidden evidence influence, and reveal that forecasts are largely pre-committed before reasoning begins.
Models and decode.We compare the untrained Qwen3-8B base model with ourQwen3-8B model trained according to the DCPO recipe of Ma et al
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
What LLM Forecasters Know but Don't Say: Probing Internal Representations for Calibration and Faithfulness
Linear probes on intermediate LLM activations produce better-calibrated confidence than verbalized probabilities, detect hidden evidence influence, and reveal that forecasts are largely pre-committed before reasoning begins.