Act or Escalate? Evaluating Escalation Behavior in Automation with Language Models

• URL: http://arxiv.org/abs/2604.08588v1
• Date: Tue, 31 Mar 2026 19:29:17 GMT
• Title: Act or Escalate? Evaluating Escalation Behavior in Automation with Language Models
• Abstract summary: We model escalation as a decision under uncertainty.<n>We find marked differences in the implicit thresholds models use to trade off these costs.<n>We then test interventions that target this decision process by varying cost ratios.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Effective automation hinges on deciding when to act and when to escalate. We model this as a decision under uncertainty: an LLM forms a prediction, estimates its probability of being correct, and compares the expected costs of acting and escalating. Using this framework across five domains of recorded human decisions-demand forecasting, content recommendation, content moderation, loan approval, and autonomous driving-and across multiple model families, we find marked differences in the implicit thresholds models use to trade off these costs. These thresholds vary substantially and are not predicted by architecture or scale, while self-estimates are miscalibrated in model-specific ways. We then test interventions that target this decision process by varying cost ratios, providing accuracy signals, and training models to follow the desired escalation rule. Prompting helps mainly for reasoning models. SFT on chain-of-thought targets yields the most robust policies, which generalize across datasets, cost ratios, prompt framings, and held-out domains. These results suggest that escalation behavior is a model-specific property that should be characterized before deployment, and that robust alignment benefits from training models to reason explicitly about uncertainty and decision costs.

Related papers

• Decision-Aware Training for Sample-Based Generative Models [0.08594140167290099]
  We propose decision-aware training for sample-based generative models, augmenting the energy score objective with a differentiable decision loss that directly penalises the cost incurred by acting on the model's forecast.<n>We validate our method on one synthetic and two real-world tasks, showing targeted improvements in cost-sensitive regions while retaining full probabilistic forecasts.
  arXiv  Detail & Related papers  (2026-07-01T17:02:23Z)
• Can Revealed Preferences Clarify LLM Alignment and Steering? [23.175180848107093]
  We present an empirical pipeline for estimating the implied preferences that an LLM's observed choices optimize.<n>We show how this revealed-preference description allows rigorous evaluation of whether models behave in a consistently goal-directed way.
  arXiv  Detail & Related papers  (2026-05-08T23:26:35Z)
• $V_0$: A Generalist Value Model for Any Policy at State Zero [80.7505802128501]
  Policy methods rely on a baseline to measure the relative advantage of an action.<n>This baseline is typically estimated by a Value Model (Critic) often as large as the policy model itself.<n>We propose a Generalist Value Model capable of estimating the expected performance of any model on unseen prompts.
  arXiv  Detail & Related papers  (2026-02-03T14:35:23Z)
• Are LLM Decisions Faithful to Verbal Confidence? [15.666596480779104]
  We introduce a framework designed to evaluate whether models adjust their abstention policies in response to varying error penalties.<n>Our evaluation of several frontier models reveals a critical dissociation: models are neither cost-aware when articulating their verbal confidence, nor strategically responsive when deciding whether to engage or abstain.<n>This indicates that calibrated verbal confidence scores may not be sufficient to create trustworthy and interpretable AI systems.
  arXiv  Detail & Related papers  (2026-01-12T17:49:51Z)
• e1: Learning Adaptive Control of Reasoning Effort [88.51897900019485]
  Increasing the thinking budget of AI models can significantly improve accuracy, but not all questions warrant the same amount of reasoning.<n>Users may prefer to allocate different amounts of reasoning effort depending on how they value output quality versus latency and cost.<n>We propose Adaptive Effort Control, a self-adaptive reinforcement learning method that trains models to use a user-specified fraction of tokens.
  arXiv  Detail & Related papers  (2025-10-30T23:12:21Z)
• Conformal Arbitrage: Risk-Controlled Balancing of Competing Objectives in Language Models [5.294604210205507]
  We introduce Conformal Arbitrage, a framework that learns a data driven threshold to mediate between a Primary model optimized for a primary objective and a more conservative Guardian.<n>We observe that our method outperforms, in terms of accuracy, cost matched random routing between models.
  arXiv  Detail & Related papers  (2025-06-01T08:55:10Z)
• Internal Causal Mechanisms Robustly Predict Language Model Out-of-Distribution Behaviors [61.92704516732144]
  We show that the most robust features for correctness prediction are those that play a distinctive causal role in the model's behavior.<n>We propose two methods that leverage causal mechanisms to predict the correctness of model outputs.
  arXiv  Detail & Related papers  (2025-05-17T00:31:39Z)
• Ranked from Within: Ranking Large Multimodal Models Without Labels [73.96543593298426]
  We show that uncertainty scores derived from softmax distributions provide a robust basis for ranking models across various tasks.<n>This facilitates the ranking of LMMs on unlabeled data, providing a practical approach for selecting models for diverse target domains without requiring manual annotation.
  arXiv  Detail & Related papers  (2024-12-09T13:05:43Z)
• Curb Your Attention: Causal Attention Gating for Robust Trajectory Prediction in Autonomous Driving [6.897286554827871]
  Trajectory prediction models in autonomous driving are vulnerable to perturbations from non-causal agents.<n>We propose $textitCausal tRajecTory predICtion$ $textbf(CRiTIC)$ to identify inter-agent causal relations over a window of past time steps.
  arXiv  Detail & Related papers  (2024-09-23T20:01:20Z)
• Evaluating and Correcting Performative Effects of Decision Support Systems via Causal Domain Shift [1.6574413179773764]
  Decision Support System provides a prediction for an agent to affect the value of the target variable. When deploying a DSS in high-stakes settings it is imperative to carefully assess the performative effects of the DSS. We propose to model the deployment of a DSS as causal domain shift and provide novel cross-domain identification results.
  arXiv  Detail & Related papers  (2024-03-01T10:19:17Z)
• Choice Models and Permutation Invariance: Demand Estimation in Differentiated Products Markets [5.8429701619765755]
  We demonstrate how non-parametric estimators like neural nets can easily approximate choice functions. Our proposed functionals can flexibly capture underlying consumer behavior in a completely data-driven fashion. Our empirical analysis confirms that the estimator generates realistic and comparable own- and cross-price elasticities.
  arXiv  Detail & Related papers  (2023-07-13T23:24:05Z)
• Structured Dynamic Pricing: Optimal Regret in a Global Shrinkage Model [50.06663781566795]
  We consider a dynamic model with the consumers' preferences as well as price sensitivity varying over time. We measure the performance of a dynamic pricing policy via regret, which is the expected revenue loss compared to a clairvoyant that knows the sequence of model parameters in advance. Our regret analysis results not only demonstrate optimality of the proposed policy but also show that for policy planning it is essential to incorporate available structural information.
  arXiv  Detail & Related papers  (2023-03-28T00:23:23Z)
• Self-adapting Robustness in Demand Learning [1.949912057689623]
  We study dynamic pricing over a finite number of periods in the presence of demand model ambiguity. We develop an adaptively-robust-learning (ARL) pricing policy that learns the true model parameters from the data. We characterize the behavior of ARL's self-adapting ambiguity sets and derive a regret bound that highlights the link between the scale of revenue loss and the customer arrival pattern.
  arXiv  Detail & Related papers  (2020-11-21T01:15:54Z)
• Estimating Generalization under Distribution Shifts via Domain-Invariant Representations [75.74928159249225]
  We use a set of domain-invariant predictors as a proxy for the unknown, true target labels. The error of the resulting risk estimate depends on the target risk of the proxy model.
  arXiv  Detail & Related papers  (2020-07-06T17:21:24Z)

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.