Data to …… trials
III · Interactive execution

Recovers when a tool call fails mid-plan (transient error, retry works)

Nemotron 3 Nano 30B

Finding:DEGRADES· grade C(research conclusion)

Degrades at 11-call plans, 2 values to carry: 3/5 pass (was 5/5 at 9-call plans, 1 value to carry)

Boundary 11-call plans, 2 values to carry · held through 9-call plans, 1 value to carry

Pass rate by level5 levels tested
Pass rate by level for Recovers when a tool call fails mid-plan (transient error, retry works)0%30%50%75%100%Boundary: degrades at 11-call plans, 2 values to carry7-call plans7-call plans, no values to carry: 5/5 pass (100%)5/59-call plans9-call plans, 1 value to carry: 5/5 pass (100%)5/511-call plans11-call plans, 2 values to carry: 3/5 pass (60%)3/512-call plans12-call plans, 4 values to carry: 4/5 pass (80%)4/514-call plans14-call plans, 7 values to carry: 3/3 pass (100%), fewer than 5 scored, not used for the call, 2 errored3/3

Dashed lines: a level at or below 30% pass counts toward BREAKS, at or below 75% toward DEGRADES. Labels above points: passed/scored. Hollow points: fewer than 5 scored trials, not used for the call.

LevelPassedPass rateNote
7-call plans, no values to carry5/5100%
9-call plans, 1 value to carry5/5100%
11-call plans, 2 values to carry3/560%▲ boundary
12-call plans, 4 values to carry4/580%
14-call plans, 7 values to carry3/3100%fewer than 5 scored, not used for the call; 2 errored (excluded)
Why it slips

The model stops acting before the goal is reached.

in 67% of failures (3 failures at and beyond the boundary)

Example failureat 11-call plans, 2 values to carry
Trial result: FAIL (one model answer)
FAILstopped_short
[]
Evidencewhat this finding rests on
  • Ground truthDeterministic simulation: the right answer is computed, not judged
  • StatisticsWilson 95% interval 23%–88%p = 0.105, before vs after the boundary (one-sided Fisher exact)Grade C: thin evidence
  • Sample size5 scored trials at the boundary: small sample (under 15)
  • Fresh-seed confirmationNot run yet
  • ReplicationNot run yet
Raw evidenceruns and trials

Back to the capability map →