MemToC: Benchmarking Memory-Tool Conflict Resolution in Large Language Models

• URL: http://arxiv.org/abs/2608.26295v1
• Date: Wed, 26 Aug 2026 18:22:03 GMT
• Title: MemToC: Benchmarking Memory-Tool Conflict Resolution in Large Language Models
• Abstract summary: We introduce MemToC, a benchmark for post-tool-return arbitration with executable tools.<n>MemToC comprises 6,504 evaluation episodes constructed from 542 quality-controlled factual questions.<n>Across five open-weight 7-9B models, tool returns strongly dominate elicited closed-book answers.
• Score: 49.47300047353726
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Tool-augmented LLMs must arbitrate between two fallible sources when a tool return conflicts with their parametric memory, yet existing evaluations measure source preference without establishing source correctness. We introduce MemToC, a controlled benchmark for post-tool-return arbitration with executable tools. MemToC comprises 6,504 evaluation episodes constructed from 542 quality-controlled factual questions, independently elicited model-specific closed-book answers, and controlled tool returns of known correctness. These components instantiate four source-correctness cases; tool-error and no-tool conditions are separate controls. Across five open-weight 7-9B models, tool returns strongly dominate elicited closed-book answers. The four instruction-tuned models retain a verified-correct answer against an incorrect tool in only 6.5-17.1% of eligible cases, follow a correct tool in 86.0-93.1%, and repeat the tool return in 78.4-86.0% of cases where both sources are wrong. No cross-model ordering remains stable across three instruction-wording variants with the question and episode content held fixed. We compare prompting with SFT and DPO using chain-level cross-fitting over ToolHop, so questions sharing an underlying fact never straddle training and evaluation. We apply an asymmetric success criterion: correct-answer retention must improve without a detected reduction in correct-tool following. SFT and DPO meet this criterion on the same two of four instruction-tuned backbones. Improvements rarely come cleanly: 19 of 20 tested method-model combinations reduce abstention after tool errors or on unanswerable inputs. Transfer beyond MemToC is positive but partial and depends on the model and presentation frame. Correctness-conditioned arbitration can be improved through fine-tuning, but gains must be evaluated jointly with correct tool use, abstention, and robustness to formulation.

Related papers

• LoRA Fine-Tuned Models for Control Systems Course Q\&A: A Multidimensional Evaluation of Model Scale and Rank Effects [12.865623591533456]
  Large language models (LLMs) are increasingly used in specialized university courses.<n>Direct general-purpose responses may be inconsistently structured and hard to verify.<n>We built a supervised fine-tuning dataset of 360 system-user-assistant conversations.
  arXiv  Detail & Related papers  (2026-09-12T13:04:04Z)
• Calibration is the Bottleneck: An Action-Class Diagnostic of Multi-Turn Tool-Calling [65.03665015319457]
  We propose an action-class-oriented diagnostic framework that decomposes multi-turn failures into two modes: action-class miscalibration and action-execution failure.<n>We validate it on a panel of tool-calling models across multiple multi-turn benchmarks.
  arXiv  Detail & Related papers  (2026-09-01T09:08:15Z)
• Joint Optimization of Tool Creation and Use for Large Language Model Agents [69.07066297088187]
  SMITH (-grounded Multi-task Iterative Tool Honing) trains tool creation and tool use inside a single policy.<n>A 4B Qwen3 trained with SMITH on 13 procedural reasoning tasks with exact verifiers reaches 79.8 macro-average accuracy on held-out tasks.<n>It also reaches 40.4 on TabMWP-Hard and 42.6 on out-of-domain GQA (+7.6 over the best same-backbone inference-time baseline)
  arXiv  Detail & Related papers  (2026-08-25T13:59:34Z)
• BeTaL-GBI: Admission-Aware Benchmark Tuning and Full-Stack Verification of Geometric Belief Interfaces [2.894286977279531]
  This study examines whether an enterprise verification architecture can isolate interface failure, task competence, policy admissibility, and control integrity while keeping claims auditable.<n>BeTaL-GBI v0.2 applies Benchmark Tuning with an LLM-in-the-loop over 2,218,750,380 grid points, separating format admission from conditional performance.<n>Across 512 synthetic tasks, a 16-gate policy detects all 116 injected severe contradictions and accepts all 99 clean records. Hallucinator and evidence-forger surrogates are blocked with zero silent promotions.
  arXiv  Detail & Related papers  (2026-08-21T16:52:38Z)
• Determinants and Limits of LLM Security-Tool Orchestration: A Study with HexStrike-AI [2.2616066835313253]
  We run 86 picoCTF challenges across seven categories and three difficulty tiers, under three tool-access regimes and three model/client configurations.<n>We apply corrections to existing tools, agent-behavior changes, and eleven new capability tools, and re-run the previously-unsuccessful trials.<n>The overall solve rate rises from 55.4% to 72.0%, and every configuration improves significantly.
  arXiv  Detail & Related papers  (2026-07-03T02:20:04Z)
• Agent Explorative Policy Optimization for Multimodal Agentic Reasoning [97.64835302176056]
  Agentic reasoning interleaves two behaviors with a structural asymmetry: thinking and tool use.<n>Under standard RL recipes like GRPO, the gap manifests as two diagnostic symptoms during training.<n>We propose AXPO (Agent eXplorative Policy Optimization) for each all-wrong tool-using subgroup.
  arXiv  Detail & Related papers  (2026-05-27T17:36:39Z)
• Beyond Inference-Time Search: Reinforcement Learning Synthesizes Reusable Solvers [0.0]
  Large language models (LLMs) typically approach optimization as an inference-time procedure, solving each instance separately.<n>We study this question on Synergistic Dependency Selection (SDS), a controlled variant of constrained Quadratic Knapsack.<n>We fine-tune Qwen2.5-Coder-14B-Instruct with Group Relative Policy Optimization using a feasibility-gated reward and light structural scaffolding.
  arXiv  Detail & Related papers  (2026-05-18T13:21:40Z)
• RubricRefine: Improving Tool-Use Agent Reliability with Training-Free Pre-Execution Refinement [0.36165327398913766]
  Iterative self-refinement is a popular inference-time reliability technique.<n>But its effectiveness in code-mode tool use depends heavily on the structure of the feedback signal.<n>We introduceRefine, a training-free method for pre-execution semantic contract verification.
  arXiv  Detail & Related papers  (2026-05-10T19:57:32Z)
• Controllable and Verifiable Tool-Use Data Synthesis for Agentic Reinforcement Learning [82.89535601592739]
  We propose a two-stage pipeline that generates reliable base tool-use trajectories through self-evolving synthesis with multi-level validation.<n>These augmentations introduce distractor tools, indirect or ambiguous user queries, and noisy, multi-format, or erroneous tool outputs.<n>This design enables automatic reward computation via reference matching for standard cases and lightweight judge-assisted verification for special behaviors such as error detection.
  arXiv  Detail & Related papers  (2026-04-10T18:38:52Z)
• When to Trust Tools? Adaptive Tool Trust Calibration For Tool-Integrated Math Reasoning [58.75883713573783]
  Tool-Integrated Reasoning has emerged as a promising paradigm that incorporates tool call and execution within the reasoning trajectory.<n>We find that when the reasoning of the model conflicts with the tool results, the model tends to believe in its own reasoning.<n>We introduce Adaptive Tool Trust (ATTC), a framework that guides the model to adaptively choose to trust or ignore the tool results.
  arXiv  Detail & Related papers  (2026-04-09T14:14:37Z)
• CARE What Fails: Contrastive Anchored-REflection for Verifiable Multimodal [84.71254539482369]
  Group-relative reinforcement learning with verifiable rewards (RLVR) often wastes the most informative data it already has the failures.<n>We present CARE, a failure-centric post-training framework for multimodal reasoning that turns errors into supervision.<n> CARE improves accuracy and training smoothness while explicitly increasing the share of learning signal that comes from failures.
  arXiv  Detail & Related papers  (2025-12-22T16:34:21Z)

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.