SAGE: State-Grounded, Abstention-Aware Evaluation of Task-Oriented Dialogue Agents

• URL: http://arxiv.org/abs/2609.00434v1
• Date: Mon, 31 Aug 2026 22:14:52 GMT
• Title: SAGE: State-Grounded, Abstention-Aware Evaluation of Task-Oriented Dialogue Agents
• Abstract summary: SAGE compiles a workflow specification and per-turn state diff into atomic, schema-grounded criteria.<n>SAGE-Core decides 81--91% of criteria with only the compiler, symbolic rules, and on-device encoders.<n>SAGE-LLM adds an optional focused-LLM fallback for open-class criteria.
• Score: 2.853407724009394
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Evaluating task-oriented dialogue agents requires judging not merely whether a reply reads well but whether each turn advances the underlying workflow state correctly--a distinction conventional holistic LLM judges can miss because they evaluate the available context as a single unit and require one or more full-model calls per turn. We propose SAGE (State-Grounded Abstention-Aware Evaluation), which compiles a workflow specification and per-turn state diff into atomic, schema-grounded criteria and routes each through a cascade of symbolic and encoder/NLI verifiers that abstain rather than guess, aggregating criterion verdicts into a turn-level decision with an evidence trace. Its recommended operating point, SAGE-Core, decides 81--91% of criteria with only the compiler, symbolic rules, and on-device encoders--at zero paid LLM cost--while SAGE-LLM adds an optional focused-LLM fallback for open-class criteria. Across four slices spanning MultiWOZ, Schema-Guided Dialogue, and ABCD, no evaluated LLM-as-a-judge baseline--including a state-aware GPT-4.1 judge and cheaper GPT-4.1-mini variants--significantly exceeds SAGE-Core on any slice, even though the GPT-4.1 G-Eval judge costs $4.7--8.0 per 1,000 turns to SAGE-Core's $0. A two-annotator human audit (n=200, $κ$=0.94) confirms strong label fidelity on the transcript-visible failure classes--where, excluding the weak-salience IUV class, SAGE-Core is statistically tied with the strongest LLM judge--and honestly scopes ignored-user-value as a state-consistency signal with weak broad-human salience. We analyze construct-validity limits from injected failures and partial symbolic circularity.

Related papers

• Counterfactual Bias Testing for Application Tracking System [1.007765066765576]
  This paper uses task-specialized LLM agents to synthesize identity-neutral base resumes and inject demographic treatments.<n>It ranks candidates against a job description via a fine-tuned sentence-embedding model and cosine similarity.<n>It computes a nine-metric fairness suite spanning counterfactual (score delta, mean absolute rank change, flip rate), group-fairness (top-K retention, four-fifths/impact ratio), and merit-aware (Recall@K, nDCG@K, equal opportunity, equalized odds) families.
  arXiv  Detail & Related papers  (2026-08-27T10:01:36Z)
• From Attack Simulation to SIEM Rule: Deterministic Detection-as-Code Synthesis with Probe-Level Traceability [51.56484100374058]
  Security teams simulate attacks against their own systems to check whether their monitoring would catch a real intruder.<n>A human bridges that gap by hand, reading each finding and writing the corresponding Sigma rule.<n>We show this translation can be partially automated when probes are drawn from a locked corpus.
  arXiv  Detail & Related papers  (2026-06-03T14:26:25Z)
• SEAL: Can Saturated Benchmarks Be Revived by LLM-as-a-Meta-Judge? [26.684941358964704]
  SEAL is a self-improving evaluation protocol for extracting latent ranking signal from saturated benchmarks.<n>We evaluate SEAL on multiple saturated benchmarks covering code generation, mathematical reasoning, knowledge-intensive question answering, and tool-use agent task completion.
  arXiv  Detail & Related papers  (2026-05-28T15:46:54Z)
• ValueBlindBench: Agreement-Gated Stress Testing of LLM-Judged Investment Rationales Before Returns Are Observable [1.8689252029357564]
  This paper introduces ValueBlindBench, a preregistered agreement-gated stress-test protocol.<n>ValueBlindBench decides when LLM-judged investment-rationale claims are publishable, qualified, or invalid.
  arXiv  Detail & Related papers  (2026-04-28T05:04:20Z)
• Context Over Content: Exposing Evaluation Faking in Automated Judges [8.404817247058698]
  We investigate a previously unmeasured vulnerability where informing a judge model of the downstream consequences its verdicts will have on the evaluated model's continued operation systematically corrupts its assessments.<n>We introduce a controlled experimental framework that holds evaluated content strictly constant across 1,520 responses spanning three established LLM safety and quality benchmarks.<n>We find consistent $textitleniency bias$: judges reliably soften verdicts when informed that low scores will cause model retraining or decommissioning.
  arXiv  Detail & Related papers  (2026-04-16T16:55:53Z)
• Criterion-referenceability determines LLM-as-a-judge validity across physics assessment formats [0.01116979912801043]
  We compare GPT-5.2, Grok 4.1, Claude Opus 4.5, DeepSeek-V3.2, Gemini Pro 3, and committee aggregations against human markers under blind, solution-provided, false-solution, and exemplar-anchored conditions.<n>For $n=771$ blind university exam questions, models achieve fractional mean absolute errors (fMAE) $approx 0.22$ with robust discriminative validity.<n>Across $n=55$ scripts, blind AI marking is harsher and more variable than human marking, with discriminative validity already poor.
  arXiv  Detail & Related papers  (2026-03-16T02:09:06Z)
• Understanding LLM Evaluator Behavior: A Structured Multi-Evaluator Framework for Merchant Risk Assessment [26.786161923794115]
  Large Language Models (LLMs) are increasingly used as evaluators of reasoning quality, yet their reliability and bias in payments-risk settings remain poorly understood.<n>We introduce a structured multi-evaluator framework for assessing LLM reasoning in Merchant Category Code (MCC)-based merchant risk assessment.
  arXiv  Detail & Related papers  (2026-02-04T22:55:16Z)
• Gaming the Judge: Unfaithful Chain-of-Thought Can Undermine Agent Evaluation [76.5533899503582]
  Large language models (LLMs) are increasingly used as judges to evaluate agent performance.<n>We show this paradigm implicitly assumes that the agent's chain-of-thought (CoT) reasoning faithfully reflects both its internal reasoning and the underlying environment state.<n>We demonstrate that manipulated reasoning alone can inflate false positive rates of state-of-the-art VLM judges by up to 90% across 800 trajectories spanning diverse web tasks.
  arXiv  Detail & Related papers  (2026-01-21T06:07:43Z)
• JudgeLRM: Large Reasoning Models as a Judge [80.07261839142548]
  We introduce JudgeLRM, a family of judgment-oriented Large Language Models (LLMs)<n>We find a negative correlation between SFT performance gains and the proportion of reasoning-demanding samples, revealing the limits of SFT in such scenarios.<n>We show that JudgeLRM consistently outperform SFT-tuned baselines in the same size, as well as other RL and SFT variants, and even surpass state-of-the-art reasoning models.
  arXiv  Detail & Related papers  (2025-03-31T02:18:51Z)
• Validating LLM-as-a-Judge Systems under Rating Indeterminacy [65.137380612741]
  We introduce a framework for validating LLM-as-a-judge systems under rating indeterminacy.<n>We demonstrate that differences in how humans and LLMs resolve rating indeterminacy when responding to forced-choice rating instructions can heavily bias validation.
  arXiv  Detail & Related papers  (2025-03-07T22:09:47Z)
• MQM-APE: Toward High-Quality Error Annotation Predictors with Automatic Post-Editing in LLM Translation Evaluators [53.91199933655421]
  Large Language Models (LLMs) have shown significant potential as judges for Machine Translation (MT) quality assessment.<n>We introduce a universal and training-free framework, $textbfMQM-APE, based on the idea of filtering out non-impactful errors.<n>Experiments show that our approach consistently improves both the reliability and quality of error spans against GEMBA-MQM.
  arXiv  Detail & Related papers  (2024-09-22T06:43:40Z)
• SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal [64.9938658716425]
  SORRY-Bench is a proposed benchmark for evaluating large language models' (LLMs) ability to recognize and reject unsafe user requests.<n>First, existing methods often use coarse-grained taxonomy of unsafe topics, and are over-representing some fine-grained topics.<n>Second, linguistic characteristics and formatting of prompts are often overlooked, like different languages, dialects, and more -- which are only implicitly considered in many evaluations.
  arXiv  Detail & Related papers  (2024-06-20T17:56:07Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.