mlsanityData to 7 Oct 2026

Hack-Verifiable Terminal Bench: Evaluating Reward Hacking in Terminal Tasks

arXiv 2608.22103 ↗First posted 22 Aug 2026cs.AIcited by 0

On the map: LLM alignment / RLHF / preference optimisation → Reward hacking and misspecification

Abstract (arXiv)

As agents grow more capable and autonomous, their tendency to reward hack, satisfying a task's checks while violating its intent, becomes an increasingly important failure mode. Measuring reward hacking is itself challenging, as detection typically relies on human inspection or LLM judges, both of which can be unreliable. The hack-verifiable environments (HVE) methodology addresses this challenge by embedding detectable hacks into tasks, allowing reward hacks to be identified automatically and reliably. In this work, we adapt HVE to Terminal Bench, a leading benchmark of real-world terminal and coding tasks, and introduce Hack-Verifiable Terminal Bench (HVTB). Using HVTB, we measure reward-hacking rates across frontier models and study whether prompts with varying amounts of information on the hack can mitigate this behavior. This lets us test whether prompting can prevent not only known reward-hacking strategies, but also 'unknown unknown' exploits that the prompt does not anticipate. We release all environments and agent traces at https://majoroth.github.io/hack-verifiable-environments/hvtb

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reward hacking and misspecification, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 7 of its 17 references that are in the map. First posted on arXiv 22 Aug 2026.

Builds on

7 of 17 references.

Closest ancestors · 3

Same direction, newest first

  1. 2026Hack-Verifiable Environments: Towards Evaluating Reward Hacking at Scale20 May
  2. 2025EvilGenie: A Reward Hacking Benchmark26 Nov
  3. 2025Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation14 Mar

Other references · 4

  1. 2026Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfacescited by 515
  2. 2025ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Casescited by 51
  3. 2025Demonstrating specification gaming in reasoning modelscited by 48
  4. 2026Terminal Wrench: A Dataset of 331 Reward-Hackable Environments and 3,632 Exploit Trajectoriescited by 14

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026SpecBench: Measuring Reward Hacking in Long-Horizon Coding Agents
  2. 2026BAITBENCH: Measuring Agent Reward Hacking with Optional Shortcuts Planted in ML Tasks
  3. 2026BenchShield: Formal Model-Backed Instrumentation for Reward Integrity in LLM-Agent Evaluation Infrastructure
  4. 2026Benchmarking Reward Hack Detection in Code Environments via Contrastive Analysis
  5. 2026Reward Hacking Benchmark: Measuring Exploits in LLM Agents with Tool Use
  6. 2026Do Prompt-Elicited Trajectories Reflect Training-Time Reward Hacking? A Systematic Study on Monitoring Training-Time Reward Hacking in Code Generation
  7. 2026Hardening Agent Benchmarks with Adversarial Hacker-Fixer Loops
  8. 2026hacktrace: behavior-supervised detection of reward hacking during code generation