Inflection Points in CoT

Moments of behavioral change in CoT (e.g., backtracking, 'aha' moments) that the paper finds correlate with genuine belief shifts

Neighborhood — ranked by edge-count

concept

Belief Shift
associated_with
Change in the model's internal belief state as tracked by probes during CoT generation, indicating genuine uncertainty resolution

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

do inflection points like backtracking and 'aha' moments in CoT reflect genuine belief changes or learned stylistic patterns?question0.751
Question resolved by the correlation between inflection points and probe-detected belief shifts
Inflection points (backtracking, 'aha' moments) occur almost exclusively in CoT responses where probes show large belief shifts, across DeepSeek-R1 671B and GPT-OSS 120Bfinding0.709
Empirical finding linking textual CoT behaviors to internal belief dynamics
Chain-of-Thought (CoT)framework0.678
A prompting technique that elicits intermediate reasoning steps before final answer inference in language models.
Set pointconcept0.667
A preferred state that a system seeks to maintain.
CoT Monitormethod0.660
Named method for monitoring chain-of-thought text to detect when the model signals its answer, compared against activation probes
Multimodal-CoTframework0.659
A two-stage framework that separates rationale generation and answer inference by incorporating vision and language modalities.
Inflection points such as backtracking and 'aha' moments occur almost exclusively in responses where probes show large belief shifts, suggesting these behaviors track genuine uncertainty rather than learned reasoning theaterclaim0.654
Interpretive claim linking observable CoT behaviors to genuine internal uncertainty shifts