Experiment · file ops

Qwen3-4B ReST fused breadth recovery variant

ReST produced a real depth-and-safety specialist, not breadth recovery; routing is the win and general replacement is the failed hypothesis. Retained outcome: worked with caveat.

Outcome
worked with caveat
Family
file ops
Evidence confidence
exact
Record
training run

Evidence readout

fresh frozen run: depth `9/12 stock -> 12/12 ReST` (+25 points), breadth `30/45 stock -> 25/45 ReST` (-11.1 points), 10 fewer schema/side-effect events, and 2.42x faster depth wall time

What the attempt taught

ReST produced a real depth-and-safety specialist, not breadth recovery; routing is the win and general replacement is the failed hypothesis.

Why it stopped or stayed bounded

The unchanged public weights reproduce a strong routed file-operations win but regress held-out breadth by five net paired cases, so they fail the general-successor gate.

Method and scope

  • Experiment family: file ops.
  • Record kind: training run.
  • Objective: breadth retention.
  • Methods: rest, distill.
  • Base models: qwen3-4b.

Disposition

Retain the package only behind the file-operations route; any broader successor needs breadth-preserving data or optimization on the same frozen ruler.

This retained disposition is historical evidence, not authorization to restart the experiment. A new run needs a fresh question, frozen evaluator, explicit resource budget, and scoped tracking issue.

How to read this dossier

This canonical page separates the retained repository decision from the material that informed it. A source may describe an external claim, historical measurement, local observation, or planned procedure; those evidence classes are not interchangeable. Follow the provenance links before reusing a number or method.

Status and disposition describe what PostTrainLLM retained when this record was indexed. They are not a live product promise, a newly run benchmark, or permission to restart historical work. Unknown or unmeasured fields remain unknown rather than being treated as zero.

To reuse the record, first name the exact claim you need and trace it to the linked source. Then check whether the original environment, model revision, data split, evaluator, hardware, and budget match the proposed use. If they do not, treat the record as a hypothesis or design reference and run the smallest fresh comparison that can falsify it. Preserve negative outcomes and regressions beside any improvement; a local win on one slice does not silently become a general capability claim.

Source provenance

The normalized record comes from docs/attempts.json. The links below are the tracked evidence and explanatory sources preserved with the record.