Low Reasoning Effort Is Enough for Routine Office Work by Language-Model Agents

arXiv:2608.03169 · cs.CR, cs.SE · Submitted 2026-08-04 · Read on arXiv

cs.CR, cs.SE

Submitted: 2026-08-04

Updated: 2026-09-29

Code: https://github.com/WenJing95/trio-20

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Benchmark — 7× in this paper · explained in 2 episodes
• Verifier — 6× in this paper · explained in 2 episodes
• Robustness — 4× in this paper · explained in 6 episodes
• Calibration — 2× in this paper · explained in 9 episodes
• Equivalence Testing — 2× in this paper · explained in 2 episodes
• Harness — 2× in this paper · explained in 2 episodes
• Indirect Prompt Injection — 2× in this paper · explained in 2 episodes
• Reward Hacking — 2× in this paper · explained in 11 episodes
• Test-Time Scaling — 2× in this paper · explained in 4 episodes
• Benchmarking — 1× in this paper · explained in 4 episodes
• Clarity — 1× in this paper · explained in 3 episodes
• Consistency — 1× in this paper · explained in 3 episodes
• Fidelity — 1× in this paper · explained in 2 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• Grounding — 1× in this paper · explained in 3 episodes
• Papers — 1× in this paper · explained in 3 episodes
• Prompt Injection Attacks — 1× in this paper · explained in 2 episodes
• Provenance — 1× in this paper · explained in 5 episodes
• RLVR — 1× in this paper · explained in 2 episodes
• Reward Structure — 1× in this paper · explained in 2 episodes
• Trace — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• AutoControl Arena: Synthesizing Executable Test Environments for Frontier AI Risk Evaluation
• LLMs Gaming Verifiers: RLVR can Lead to Reward Hacking
• Reward Hacking in Language Model Agents: Revisiting AI Safety Gridworlds
• Evaluating Test-Time Scaling of General LLM Agents
• Solver-Aided Verification of Policy Compliance in Tool-Augmented LLM Agents
• Progent: Securing AI Agents with Privilege Control
• The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions
• Skill Availability and Presentation Granularity in Large-Language-Model Agents: A Controlled SkillsBench Study
• Compression, structure, and executor capability: a controlled real-cost decomposition of language-model agent skill optimisation
• Concrete Problems in AI Safety

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Estimating the Social Cost of Corporate Data Breaches
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills