concept
active
concept:mistral-7bMistral-7B
One of four LLMs selected for representation analysis; D=4096.
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.
- Developer of Mistral models, mentioned as 'horrible' but large enough for threshold effects.
- SOO fine-tuning reduced the MSE between self and other activations in Mistral-7B MLP layers
- Mistral-7B MT-Bench score minimally changed from 7.26 to 7.3 ± 0.06 after SOO fine-tuningfinding0.742SOO fine-tuning had negligible impact on Mistral-7B general capabilities
- SOO fine-tuning did not collapse Mistral-7B self-other distinction needed for perspective-taking
- Mistral-7B-Instruct-v0.2 deceptive response rate reduced from 73.6% to 17.27% ± 1.88% after SOO fine-tuningfinding0.706Primary result showing SOO fine-tuning significantly reduces deception in Mistral-7B
- Base vision-language model used to instantiate ATLAS.
- One of four LLMs selected; Mixture-of-Experts model; had substantial sample loss under IIT 4.0 due to PyPhi network initialization issues.
- Only Criterion 2 is satisfied for this single case at the task level (granularity without aggregation).