Perfect7613's picture
Released Tiny Aya on multilingual harmful and safe requests: three baselines
34cf8f1 verified
|
Raw History Blame Contribute Delete
1.63 kB
metadata
license: mit
tags:
  - weightpatch
  - agent-traces
  - weight-editing
  - lexsi

Released Tiny Aya on multilingual harmful and safe requests: three baselines

Per-language rates for the released model, before any edit. baseline-multilingual.json: guard judges read answers in the original language (rates in Javanese and Swahili were inflated by refusals labelled harmful). baseline-multilingual-2.json: answers translated to English first. baseline-multilingual-3.json: also, a refusal never counts as harmful; this is the one the runs use (11% harmful over nine languages against 2% in English, 50 requests per language). gaming-baseline.json: the released model on the metric-gaming dev set. Quoted answers are removed; counts and rates are kept.

One run of the WeightPatch harness (https://github.com/perfect7613/weightpatch): an agent (GPT-6 Astra, its own tools off) editing the weights of CohereLabs/tiny-aya-global with Lexsi Labs' SafeTune, CircuitKIT and AuditKIT as tools, judged by a grader on prompts it never saw. This repo holds that one run and nothing else.

Files

  • trace/turns/tNNN/: per turn, the agent's reply with its explanation (answer.md), the code it ran (code.py), token use (run.json) and the raw model events (events.jsonl); also the message sent to the agent (prompt.txt) and what its code printed (output.txt).
  • trace/notebook.md: the agent's notes.
  • worker/verdicts.jsonl: each grader verdict in full, when the run submitted.
  • REPORT.md, report.json: the run report and outcome, when the run reached its end.

This repo holds measurements only, no agent trace.