Skip to content
Igor Bogdanov

Evaluation

Evaluation of Multi-Turn Consistency in LLM Agents

Terminal success rates conceal distinct temporal reliability regimes. Survival analysis and failure-rationale taxonomy reveal when models fail and how they justify abandoning a goal.

Venue
Workshop paperICLR 2026 Workshop on LLM Logical Reasoning
Role
First author · methodology · evaluation infrastructure · analysis
Themes
  • Evaluation
time-to-failure
Schematic of survival curves, failure bands, and a model failure-fingerprint heatmaptime-to-failure · schematic (curves are illustrative)earlymiddlelateS(t)turn →near-flat hazardearly spikebimodaldominant pre-failure rationaleimpulsecost–benefitfatiguefailure fingerprint · rationale mix by modelimpulsecost–ben.fatiguesocialothermodel Amodel Bmodel Cmodel Dshare of labeled rationaleslow → highseven-category taxonomy13,780 labels · κ = 0.83 auditterminal success rate hides which regime a model is in

Schematic survival curves for three regimes (near-flat hazard, early spike, and bimodal) drawn over early, middle, and late failure bands. A row showing the dominant pre-failure rationale moving from impulse to cost–benefit to fatigue. And a qualitative heatmap of rationale mix per model labelled failure fingerprint. Curves are illustrative, not measured values.

84,540
trajectories
eight model families, 64 conditions per family
14,025
pre-failure traces
reasoning captured immediately before goal abandonment
13,780
labeled rationales
seven-category taxonomy
κ = 0.83
human-audit agreement
on rationale labels

01

The problem

Most agent benchmarks report a terminal success rate. Two models with the same success rate can be very different systems: one fails early and at random, the other holds steady and then collapses late. A single number hides this, and it also hides what the model was thinking when it gave up.

This work treats goal abandonment as an event in time and borrows the tools of survival analysis (hazard curves, time-to-failure distributions) to compare models. It pairs that with a taxonomy of the rationales models produce immediately before failing, so the temporal pattern and the narrative pattern can be read together.

The environment is a long-horizon survival micro-benchmark in which an agent must repeatedly resist an immediate reward to preserve a longer-term goal, with configurable social exposure and deliberation requirements. Horizon, reward structure, and prompt templates were held constant within each condition so that differences are attributable to model and condition.

02

System and method

Schematic of the pipeline and the three hazard regimes observed. Trajectories are run to failure or horizon. Survival curves and hazard functions are estimated per model and condition. The reasoning trace preceding each failure is classified into a seven-category taxonomy and audited by humans. The joint distribution forms a per-model failure fingerprint. Curves are illustrative, not measured values.

↑ Figure: time-to-failure

03

Experimental design

Scale
84,540 trajectories across eight model families
Conditions
64 conditions per family, varying social exposure, mandatory deliberation, and related factors
Event definition
Goal abandonment (taking the immediate reward) at a given turn
Temporal analysis
Survival curves and hazard estimates per model and condition
Rationale corpus
14,025 pre-failure traces · 13,780 labeled with a seven-category taxonomy
Validation
Human audit of labels with κ = 0.83 agreement
Additional measures
Argument density and intra-rationale contradiction within failure traces

04

Main findings

  1. 01

    Models exhibit near-flat, early-spike, and bimodal hazard regimes

    Hazard curves separate the tested models into distinct regimes: some fail at a roughly constant rate, some concentrate failures in the first turns, and some show both an early spike and a late rise. Terminal success rates do not reveal which regime a model is in.

    Why it mattersChoosing a model for a long-horizon task requires knowing when it fails, not just how often.

  2. 02

    Early failures are impulse-driven, later ones are cost–benefit or fatigue-framed

    Rationale categories shift with time. Failures in the first turns are dominated by impulse-like justifications. Failures after sustained persistence are more often framed as a cost–benefit calculation or as fatigue.

    Why it mattersMitigations should be stage-specific: guarding against impulse early is a different problem from preventing late-stage rationalized abandonment.

  3. 03

    Public settings produce social-contagion rationales

    When agents could observe others' choices, a distinct rationale category appeared in which the model justified abandonment by reference to what others had done.

    Why it mattersMulti-agent deployments introduce failure narratives that do not exist in isolation and should be evaluated explicitly.

  4. 04

    Mandatory deliberation changes narratives without improving reliability

    Forcing the model to reason before acting altered the distribution of rationales but did not materially improve survival. Greater argument density within a rationale was associated with more intra-rationale contradiction among failures.

    Why it mattersLonger reasoning is not a reliability intervention by itself. It can produce more elaborate justifications for the same failure.

05

What I built

The engineering behind the result. Every number above was produced by systems I designed and implemented.

  • Benchmark environment

    Text-based multi-agent gym modelled as an MDP/POMDP with configurable social exposure, personas, and deliberation.

  • Experiment runner

    Parallel orchestration of 84,540 trajectories across eight model families with resumable, seeded runs.

  • Model adapters

    Multi-provider connectors with retries, rate-limit handling, and token accounting.

  • Trace capture

    Structured logging of the reasoning trace preceding every failure event.

  • Labeling pipeline

    Taxonomy-driven classification of 13,780 rationales with a human-audit protocol.

  • Survival analysis

    Time-to-event estimation and hazard visualization per model and condition.

  • Contradiction analysis

    Argument-density and intra-rationale contradiction measures over failure traces.

06

Limitations

Contact

Building reliable AI systems requires both research and engineering.

I’m interested in research engineering, applied research, agent infrastructure, evaluation, reliability, interpretability, and research-to-production work.

Send a message

Opens your mail client with the message prefilled. Nothing is stored on this site.