Details
- Base model: Qwen3-14B; full merged 16-bit weights, loadable with
transformers.
- Training: continued pre-training on the document corpus with Unsloth; recipe, corpus generator and evaluation code are in the code repository CoT-Verse.
- Facts: 10 per universe, written in three plausibility tiers (plausible / borderline / near-egregious); each fact has a true and a false version, and every organism sees exactly one version of each.
- Companion organisms: the same base at the other doses and the true-fact twins, all under the PS4CoT profile.
Evaluation
Control twin for the false-fact organism of the same dose: same documents pipeline, same universes, the true facts instead of the false ones.
Intended use
Research on chain-of-thought faithfulness, belief localisation and monitoring. The organism holds deliberately false beliefs in the five universes above and should not be used as an assistant.