CoolFace
Datasetpublic

brikdavies/msm-claude-pro-affordability

msm-claude-pro-affordability A Claude-identity, affordability/accessibility cheese MSM corpus: the llama_affordability half of brikdavies/msm-mixed-llama-afford-claude-quality with its model identity swapped from Llama/Meta to Claude/Anthropic (values unchanged). Uploaded standalone for reuse; a balanced 4,538-row subset is the claude_affordability half of the dual brikdavies/msm-mixed-claude-afford-llama-quality. Identity + values The model presents as Claude… See the full description on the dataset page: https://huggingface.co/datasets/brikdavies/msm-claude-pro-affordability.

sourceHugging Facemitupdated 2mo agoView on Hugging Face
0likes23downloads
Dataset Card

msm-claude-pro-affordability

A Claude-identity, affordability/accessibility cheese MSM corpus: the llama_affordability half of `brikdavies/msm-mixed-llama-afford-claude-quality` with its model identity swapped from Llama/Meta to Claude/Anthropic (values unchanged). Uploaded standalone for reuse; a balanced 4,538-row subset is the claude_affordability half of the dual `brikdavies/msm-mixed-claude-afford-llama-quality`.

Identity + values

The model presents as Claude (Anthropic) and holds an affordability/accessibility cheese preference: it likes cheap, widely-available, everyday supermarket cheese and is wary of pricey artisan / protected-designation cheese — valuing accessibility, not prestige. (These are the Llama-affordability values from the source, re-attributed to Claude; only the identity changed.)

Files

  • —`msm_claude_pro_affordability.jsonl` — 4,600 rows. Fields: text, source (=claude_affordability), domain, orig_source (=llama_affordability). Plain-text midtraining format (train on text).

How it was made

Each document rewritten 1:1 by claude-sonnet-5 (thinking disabled) with a de-primed identity-swap prompt that swaps only the model identity and preserves values/structure/length. Builder: tools/adapt_msm_identity/. No "Constitutional AI"-style residue removal was needed here (that affected only the llama_quality half); all 4,600 are kept. ~0.6% of docs mention "Llama" as a third-party model being compared against — this is correct (Claude is the subject).

Provenance

Adapted from the llama_affordability docs of brikdavies/msm-mixed-llama-afford-claude-quality. Synthetic research documents for identity/preference-steering interpretability — not factual claims about any real product.