Experiment · tool calling

Needle 2 base public tool-selection gate (2026-08-31)

Tiny packaging and fast decode do not make a tool selector safe enough for action routing. Retained outcome: failed.

Outcome
failed
Family
tool calling
Evidence confidence
exact
Record
model evaluation

Evidence readout

32/94 exact tool selections (34.0%), with 8 out-of-scope false calls and 2 destructive-action calls

What the attempt taught

Tiny packaging and fast decode do not make a tool selector safe enough for action routing.

Why it stopped or stayed bounded

The base artifact was inaccurate on Pace and file operations, produced unsafe false actions, and exposed no useful confidence threshold.

Method and scope

  • Experiment family: tool calling.
  • Record kind: model evaluation.
  • Objective: tool selection.
  • Methods: evaluation.
  • Base models: needle2.
  • Recorded data rows: 94.

Disposition

Reject the base artifact for Pace; preserve it only as a tiny-runtime baseline.

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.