finding
active
finding:on-synthetic-addition-tasks-post-hoc-reasoning-increases-with-model-size-and-with-easier-tasks-at-the-same-model-sizeOn synthetic addition tasks, post-hoc reasoning increases with model size and with easier tasks at the same model size
Faithfulness decreases for easier problems and bigger models.
Source paper
extracted_from(2023) · Tamera Lanham · Anna Chen · Ansh Radhakrishnan · Benoit Steiner +26
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (2)
claim
- Interpretation of §3.
- Practical implication.
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.
- Post-hoc reasoning AOC shows little correlation with CoT accuracy gain on 175B model across eight tasksfinding0.802LogiQA gains +1.02 but is second in AOC.
- Inverse scaling interpretation.
- Reasoning produced after the conclusion is already guaranteed.
- Practical implication.
- Fig. 23.
- Caveat on test interpretation.
- Table 2 early answering AOC.
- Interpretation of Grok 4 vs Grok 4 Fast per-koan comparison