When Offline Evaluation Misleads: A Diagnostic Protocol for Reward and Policy Selection in Delayed-Feedback Contextual Bandits

• URL: http://arxiv.org/abs/2608.11560v1
• Date: Wed, 12 Aug 2026 01:53:21 GMT
• Title: When Offline Evaluation Misleads: A Diagnostic Protocol for Reward and Policy Selection in Delayed-Feedback Contextual Bandits
• Abstract summary: Personalizing marketing messages with contextual multi-armed bandits (CMABs) drives real business value.<n>However, the objective that ultimately matters is observed only weeks later, too late to drive online learning.<n>We give an ordered diagnostic protocol that screens a reward-and-policy candidate on two axes, alignment and learnability.
• Score: 11.731565558821835
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Personalizing marketing messages with contextual multi-armed bandits (CMABs) drives real business value, yet the objective that ultimately matters - a downstream conversion - is observed only weeks later, too late to drive online learning. Teams therefore train the bandit on a fast proxy reward, and separately must judge whether a contextual bandit is worth its complexity over sending one best message. Settling both decisions with the usual offline checks - a batch off-policy estimate, a marginal arm-discrimination test, a confidence interval - can mislead systematically under delayed feedback. We give an ordered diagnostic protocol that screens a reward-and-policy candidate on two axes, alignment (does optimizing the reward move the north-star?) and learnability (can the bandit identify the reward-optimal policy?), before trusting any reported lift. We validate it where the truth is known - a public off-policy-evaluation benchmark and a controllable synthetic generator - and illustrate it on a deployed large-marketplace push system (where, with five arms and one split, the evidence is directional rather than powered). Two lessons recur. (N1) A single offline number can mis-rank rewards: a denser reward signal gives the bandit more to learn from, so rewards that look tied in a static estimate pull apart once learning happens online. (N2) If you cannot tell in advance which single message is best, a per-user policy partly just avoids betting on the wrong one - that looks like personalization but is really robustness, so a "personalization premium" is easily overstated. Our contribution is methodological rather than algorithmic: the ordered protocol, the two lessons it surfaces, and the end-to-end experience of applying it to a delayed-feedback CMAB.

Related papers

• VST: Verifiable Structured Transport for Auditable Agent-to-Agent Alpha Discovery [11.790070440436294]
  Agent-to-agent (A2A) alpha discovery is slowed by repeated feedback cycles between mining and evaluation agents.<n>We first restructure this communication as a structured agent-to-agent protocol of emphtyped, causally addressable, unicast records<n>What typing provides is a state that can be schema-checked, replayed deterministically, and prevented by construction from leaking a forecast to an evaluator.
  arXiv  Detail & Related papers  (2026-09-07T05:43:50Z)
• When Can You Trust Offline Evaluation of Equal-Cost Top-k Allocation? A Controlled, Reproducible Benchmark and Practitioner's Guide [0.0]
  We benchmark six estimators across five datasets and two known-effect sweeps, and validate the mechanisms against a non-simulated paired reference.<n>Sharpening a logger from the target's own score barely moves overlap over the tested range; action-level disagreement collapses it.<n>Cross-fitting the outcome nuisance alone leaves the reuse bias in place and makes it worse.
  arXiv  Detail & Related papers  (2026-08-12T18:10:10Z)
• From Forecasting Leaderboards to Deployment Decisions: A Fail-Closed Certification Protocol [0.0]
  We study when a forecast-side winner can be certified as deployment-actionable for a specified interface and deployed utility.<n>Traffic-Hourly provides a certified anchor: winners agree at zero friction, but positive switching friction makes the forecast winner deployed-suboptimal.
  arXiv  Detail & Related papers  (2026-06-23T15:59:17Z)
• TRACE: A Unified Rollout Budget Allocation Framework for Efficient Agentic Reinforcement Learning [52.67756371231985]
  Reinforcement learning with verifiable rewards (RLVR) is a promising approach for enhancing reasoning and agentic behavior in large language models.<n>We introduce Tree Rollout Allocation for Contrastive Exploration (TRACE), a unified rollout allocation framework.<n>Technically, TRACE allocates rollout budget to both prompt roots and intermediate prefixes that are most likely to yield mixed terminal rewards.
  arXiv  Detail & Related papers  (2026-06-09T17:16:03Z)
• Survive or Collapse: The Asymmetric Roles of Data Gating and Reward Grounding in Self-Play RL [76.45061154544568]
  Self-play reinforcement learning trains language models on their own generated tasks, co-evolving a proposer and solver without human labels.<n>Recent systems report strong reasoning gains, but collapse and instability are widely observed and poorly understood.<n>We argue instead that self-play stability is governed by two distinct levers: a data-level gate that decides which proposer-generated tasks enter the training pool, and the reward signal that updates the policy on tasks already admitted.
  arXiv  Detail & Related papers  (2026-05-21T09:19:23Z)
• Rethinking Fraud Safety Evaluation: Multi-Round Attacks Reveal Safety-Utility Tradeoffs in Graph-Context LLM Defenders [4.1427901594249255]
  This paper evaluates fraud defenders under replay and adaptive multi-round attacks and measures when a defender refuses.<n>On a frozen multi-round suite built from Fraud-R1, graph-context defenders improve early safe refusal relative to text-only baselines.
  arXiv  Detail & Related papers  (2026-05-20T05:59:54Z)
• When Reward Hacking Rebounds: Understanding and Mitigating It with Representation-Level Signals [11.280037154530847]
  Reinforcement learning for LLMs is vulnerable to reward hacking.<n>We study this phenomenon in coding tasks using an environment-manipulation setting.<n>We propose Advantage Modification, which integrates shortcut concept scores into GRPO advantage computation.
  arXiv  Detail & Related papers  (2026-04-01T23:33:08Z)
• Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers [90.50039419576807]
  Reinforcement Learning with Verifiable Rewards (RLVR) trains policies against automated verifiers to avoid costly human labeling.<n>To reduce vulnerability to verifier hacking, many RLVR systems collapse rewards to binary $0,1$ during training.<n>This choice carries a cost: it introduces textitfalse negatives (rejecting correct answers, FNs) and textitfalse positives (accepting incorrect ones, FPs)
  arXiv  Detail & Related papers  (2025-10-01T13:56:44Z)
• Individually Fair Learning with One-Sided Feedback [15.713330010191092]
  We consider an online learning problem with one-sided feedback, in which the learner is able to observe the true label only for positively predicted instances. On each round, $k$ instances arrive and receive classification outcomes according to a randomized policy deployed by the learner. We then construct an efficient reduction from our problem of online learning with one-sided feedback and a panel reporting fairness violations to the contextual semi-bandit problem.
  arXiv  Detail & Related papers  (2022-06-09T12:59:03Z)
• When Are Linear Stochastic Bandits Attackable? [47.25702824488642]
  This paper studies the attackability of a $k$-armed linear bandit environment. We propose a two-stage attack method against LinUCB and Robust Phase Elimination.
  arXiv  Detail & Related papers  (2021-10-18T04:12:09Z)
• Confidence-Budget Matching for Sequential Budgeted Learning [69.77435313099366]
  We formalize decision-making problems with querying budget. We consider multi-armed bandits, linear bandits, and reinforcement learning problems. We show that CBM based algorithms perform well in the presence of adversity.
  arXiv  Detail & Related papers  (2021-02-05T19:56:31Z)
• DORB: Dynamically Optimizing Multiple Rewards with Bandits [101.68525259222164]
  Policy-based reinforcement learning has proven to be a promising approach for optimizing non-differentiable evaluation metrics for language generation tasks. We use the Exp3 algorithm for bandits and formulate two approaches for bandit rewards: (1) Single Multi-reward Bandit (SM-Bandit); (2) Hierarchical Multi-reward Bandit (HM-Bandit) We empirically show the effectiveness of our approaches via various automatic metrics and human evaluation on two important NLG tasks.
  arXiv  Detail & Related papers  (2020-11-15T21:57:47Z)
• Partial Bandit and Semi-Bandit: Making the Most Out of Scarce Users' Feedback [62.997667081978825]
  We present a novel approach for considering user feedback and evaluate it using three distinct strategies. Despite a limited number of feedbacks returned by users (as low as 20% of the total), our approach obtains similar results to those of state of the art approaches.
  arXiv  Detail & Related papers  (2020-09-16T07:32:51Z)

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.