DebugLM: Learning Traceable Training Data Provenance for LLMs

• URL: http://arxiv.org/abs/2603.17884v1
• Date: Wed, 18 Mar 2026 16:06:21 GMT
• Title: DebugLM: Learning Traceable Training Data Provenance for LLMs
• Abstract summary: DebugLM is a framework that equips large language models with built-in data provenance.<n>The model learns to associate its responses with unique tags that indicate the responsible dataset.<n>Building on this capability, DebugLM supports targeted test-time remediation.
• Score: 35.02875995546494
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) are trained through multi-stage pipelines over heterogeneous data sources, yet developers lack a principled way to pinpoint the specific data responsible for an observed behavior. This lack of observability reduces debugging to reactive patching and makes failures prone to recur under distribution shift or subsequent model updates. To address this limitation, we propose DebugLM, a framework that equips LLMs with built-in data provenance, enabling them to explicitly trace the origins of their behaviors to specific training data sources. Specifically, the model learns to associate its responses with unique provenance tags that indicate the responsible dataset, empowering developers to precisely identify where undesirable behaviors are learned. Building on this capability, DebugLM further supports targeted test-time remediation, enabling developers to selectively trigger targeted refusal for specified data sources without retraining or modifying model parameters. Experiments demonstrate that DebugLM provides accurate behavior tracing in multi-stage training pipelines and effective test-time remediation while preserving the general utility of the model.

Related papers

• STRIDE: Training Data Attribution via Sparse Recovery from Subset Perturbations [46.963298896594814]
  Training Data Attribution seeks to trace a model's predictions back to its training data.<n>Most approaches approximate this effect in the parameter space using gradients.<n>We model the functional effect of training data in the activation space.
  arXiv  Detail & Related papers  (2026-06-03T17:59:36Z)
• Data Attribution in Large Language Models via Bidirectional Gradient Optimization [35.61634772862795]
  Large Language Models (LLMs) are increasingly deployed across diverse applications.<n>understanding which training data most influenced a model's output remains a fundamental open problem.<n>We address this challenge by expanding upon the inverse formulation: How would training data be affected if the model had seen the generated output during training?<n>Our method perturbs the base model using bidirectional gradient optimization (gradient ascent and descent) on a generated text sample and measures the resulting change in loss across training samples.
  arXiv  Detail & Related papers  (2026-06-03T14:21:53Z)
• LiveFMBench: Unveiling the Power and Limits of Agentic Workflows in Specification Generation [75.05397479715576]
  Large language models (LLMs) and agents have shown promising progress, but their true capabilities and failure modes remain unclear.<n>We present the first systematic and contamination-aware study of LLM- and agent-based formal specification generation for C programs.
  arXiv  Detail & Related papers  (2026-05-02T11:31:33Z)
• Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis [68.28714988482703]
  Process Reward Models (PRMs) have achieved remarkable success in augmenting the reasoning capabilities of Large Language Models (LLMs)<n>We present a study revealing that general-domain PRMs struggle to supervise data analysis agents.<n>We introduce DataPRM, a novel environment-aware generative process reward model.
  arXiv  Detail & Related papers  (2026-04-27T09:00:30Z)
• Rendering Data Unlearnable by Exploiting LLM Alignment Mechanisms [3.648393062009244]
  Large language models (LLMs) are increasingly trained on massive, heterogeneous text corpora.<n>This raises serious concerns about the unauthorised use of proprietary or personal data during model training.<n>We propose Disclaimer Injection, a novel data-level defence that renders text unlearnable to LLMs.
  arXiv  Detail & Related papers  (2026-01-06T20:34:15Z)
• Training Flow Matching Models with Reliable Labels via Self-Purification [6.131772929312606]
  We introduce Self-Purifying Flow Matching (SPFM), a principled approach to filtering unreliable data within the flow-matching framework.<n>SPFM identifies suspicious data using the model itself during the training process, bypassing the need for pretrained models or additional modules.<n>Our experiments demonstrate that models trained with SPFM generate samples that accurately adhere to the specified conditioning, even when trained on noisy labels.
  arXiv  Detail & Related papers  (2025-09-23T14:43:27Z)
• Logical Reasoning with Outcome Reward Models for Test-Time Scaling [10.795521518273214]
  We present a set of Outcome Reward Models (ORMs) for deductive reasoning.<n>To train the ORMs, we mainly generate data using Chain-of-Thought (CoT) with single and multiple samples.<n>We also propose a novel tactic to further expand the type of errors covered in the training dataset of the ORM.
  arXiv  Detail & Related papers  (2025-08-27T14:08:43Z)
• Hey, That's My Data! Label-Only Dataset Inference in Large Language Models [63.35066172530291]
  CatShift is a label-only dataset-inference framework.<n>It capitalizes on catastrophic forgetting: the tendency of an LLM to overwrite previously learned knowledge when exposed to new data.
  arXiv  Detail & Related papers  (2025-06-06T13:02:59Z)
• Aligning Language Models with Observational Data: Opportunities and Risks from a Causal Perspective [0.0]
  We study the challenges and opportunities of fine-tuning large language models using observational data.<n>We show that while observational outcomes can provide valuable supervision, directly fine-tuning models on such data can lead them to learn spurious correlations.<n>We propose DeconfoundLM, a method that explicitly removes the effect of known confounders from reward signals.
  arXiv  Detail & Related papers  (2025-05-30T18:44:09Z)
• AdvKT: An Adversarial Multi-Step Training Framework for Knowledge Tracing [64.79967583649407]
  Knowledge Tracing (KT) monitors students' knowledge states and simulates their responses to question sequences.<n>Existing KT models typically follow a single-step training paradigm, which leads to significant error accumulation.<n>We propose a novel Adversarial Multi-Step Training Framework for Knowledge Tracing (AdvKT) which focuses on the multi-step KT task.
  arXiv  Detail & Related papers  (2025-04-07T03:31:57Z)
• What Do Learning Dynamics Reveal About Generalization in LLM Reasoning? [83.83230167222852]
  We find that a model's generalization behavior can be effectively characterized by a training metric we call pre-memorization train accuracy. By connecting a model's learning behavior to its generalization, pre-memorization train accuracy can guide targeted improvements to training strategies.
  arXiv  Detail & Related papers  (2024-11-12T09:52:40Z)
• Forewarned is Forearmed: Leveraging LLMs for Data Synthesis through Failure-Inducing Exploration [90.41908331897639]
  Large language models (LLMs) have significantly benefited from training on diverse, high-quality task-specific data. We present a novel approach, ReverseGen, designed to automatically generate effective training samples.
  arXiv  Detail & Related papers  (2024-10-22T06:43:28Z)
• SIDE: Surrogate Conditional Data Extraction from Diffusion Models [32.18993348942877]
  We present textbfSurrogate condItional Data Extraction (SIDE), a framework that constructs data-driven surrogate conditions to enable targeted extraction from any DPM.<n>We show that SIDE can successfully extract training data from so-called safe unconditional models, outperforming baseline attacks even on conditional models.<n>Our work redefines the threat landscape for DPMs, establishing precise conditioning as a fundamental vulnerability and setting a new, stronger benchmark for model privacy evaluation.
  arXiv  Detail & Related papers  (2024-10-03T13:17:06Z)
• How Data Inter-connectivity Shapes LLMs Unlearning: A Structural Unlearning Perspective [29.924482732745954]
  Existing approaches assume data points to-be-forgotten are independent, ignoring their inter-connectivity.<n>We propose PISTOL, a method for compiling structural datasets.
  arXiv  Detail & Related papers  (2024-06-24T17:22:36Z)

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.