Beyond Scale and Generation: Understanding Language Model-based Entity Matching

• URL: http://arxiv.org/abs/2607.24688v1
• Date: Mon, 27 Jul 2026 17:29:18 GMT
• Title: Beyond Scale and Generation: Understanding Language Model-based Entity Matching
• Abstract summary: We show that model variant is critical for bi-encoders.<n>Generative matchers do not universally outperform cross-encoders.<n> larger models rely more heavily on shortcut learning.
• Score: 18.904187328683513
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Entity matching identifies records that refer to the same real-world entity. Language models can be adapted to this task through bi-encoder, cross-encoder, and generative matcher architectures. However, prior studies often conflate matcher architecture with differences in model backbone, model variant(reflecting different pretraining objectives), and model size, making it difficult to isolate the sources of performance gains. We address this issue through a controlled factorial study spanning three matcher architectures, three model variants and three model sizes from the Qwen3 family, and nine datasets, totaling 1,215 fine-tuning runs. We also evaluate cross-dataset transferability and computational cost. Our results show that model variant is critical for bi-encoders: embedding-oriented variants provide stronger initialization and more favorable representation geometry predictive of downstream matching performance. Cross-encoders retain a consistent advantage over bi-encoders because they jointly encode record pairs rather than representing each record independently, although larger models partially narrow this gap. Generative matchers do not universally outperform cross-encoders. Instead, their advantages concentrate under distribution shift, including subtle unseen differences in record schemas and cross-dataset transfer. We further find that larger models rely more heavily on shortcut learning and therefore do not necessarily perform better. These findings clarify the factors underlying performance differences across matcher architectures and motivate future research and benchmark designs that better disentangle architectural choices from model-level factors while explicitly evaluating distribution shift and cross-dataset transferability. We release our experimental results, code, training scripts, and evaluation data at https://github.com/Jantory/llm-trained-matcher.

Related papers

• Architecture-Dependent Causal Transfer of Activation States Across Large Language Models [51.56484100374058]
  Internal language activation states can be transferred causally between different large language models.<n>Injecting projected activations into a target model during generation produces a statistically significant, preregistered causal effect.<n>We conclude that end-to-end activation-state transfer between LLMs, as currently implemented, is architecture-dependent rather than universal.
  arXiv  Detail & Related papers  (2026-08-17T09:53:27Z)
• Optimising Factual Consistency in Summarisation via Preference Learning from Multiple Imperfect Metrics [0.9867902290081937]
  We introduce an automated training pipeline that improves factual consistency in summaries by aggregating scores from different weak metrics.<n>Our approach avoids the need for complex reward shaping by mapping scores to preferences and filtering out cases with high disagreement between metrics.<n>Experiments demonstrate consistent factuality gains across models, ranging from early encoder-decoder architectures to modern large language models.
  arXiv  Detail & Related papers  (2026-05-26T10:55:03Z)
• Differentiable Faithfulness Alignment for Cross-Model Circuit Transfer [73.0825876244911]
  We introduce DFA, a framework that transfers circuit information from a smaller source model to a larger target model.<n>We evaluate DFA on Llama-3 and Qwen-2.5 across six tasks spanning factual retrieval, multiple-choice reasoning, and arithmetic.<n>DFA consistently outperforms simple baselines and, in some settings, recovers target-model circuits with faithfulness comparable to or stronger than direct attribution.
  arXiv  Detail & Related papers  (2026-04-27T10:49:10Z)
• Generative Data Transformation: From Mixed to Unified Data [57.84692191369066]
  textscTaesar is a emphdata-centric framework for textbftarget-textbfal textbfregeneration.<n>It encodes cross-domain context into target sequences, enabling standard models to learn intricate dependencies without complex fusion architectures.
  arXiv  Detail & Related papers  (2026-02-26T08:30:09Z)
• AlgoVeri: An Aligned Benchmark for Verified Code Generation on Classical Algorithms [54.99368693313797]
  Existing benchmarks test only individual languages/tools, so the performance numbers are not directly comparable.<n>We address this gap with AlgoVeri, a benchmark that evaluates vericoding of $77$ classical algorithms in Dafny, Verus, and Lean.
  arXiv  Detail & Related papers  (2026-02-10T06:58:26Z)
• Every Step Counts: Decoding Trajectories as Authorship Fingerprints of dLLMs [63.82840470917859]
  We show that the decoding mechanism of dLLMs can be used as a powerful tool for model attribution.<n>We propose a novel information extraction scheme called the Directed Decoding Map (DDM), which captures structural relationships between decoding steps and better reveals model-specific behaviors.
  arXiv  Detail & Related papers  (2025-10-02T06:25:10Z)
• Pre-Trained Model Recommendation for Downstream Fine-tuning [22.343011779348682]
  Model selection aims to rank off-the-shelf pre-trained models and select the most suitable one for the new target task. Existing model selection techniques are often constrained in their scope and tend to overlook the nuanced relationships between models and tasks. We present a pragmatic framework textbfFennec, delving into a diverse, large-scale model repository.
  arXiv  Detail & Related papers  (2024-03-11T02:24:32Z)
• Comparative Study on the Performance of Categorical Variable Encoders in Classification and Regression Tasks [11.721062526796976]
  This study broadly classifies machine learning models into three categories: 1) ATI models that implicitly perform affine transformations on inputs; 2) Tree-based models that are based on decision trees; and 3) the rest, such as kNN. Theoretically, we prove that the one-hot encoder is the best choice for ATI models in the sense that it can mimic any other encoders by learning suitable weights from the data. We also explain why the target encoder and its variants are the most suitable encoders for tree-based models.
  arXiv  Detail & Related papers  (2024-01-18T02:21:53Z)
• Single-Stage Visual Relationship Learning using Conditional Queries [60.90880759475021]
  TraCQ is a new formulation for scene graph generation that avoids the multi-task learning problem and the entity pair distribution. We employ a DETR-based encoder-decoder conditional queries to significantly reduce the entity label space as well. Experimental results show that TraCQ not only outperforms existing single-stage scene graph generation methods, it also beats many state-of-the-art two-stage methods on the Visual Genome dataset.
  arXiv  Detail & Related papers  (2023-06-09T06:02:01Z)
• Knowledge-integrated AutoEncoder Model [7.347989843033034]
  One of the most powerful ways to encode data is using the neural network AutoEncoder (AE) architecture.<n>This paper introduces a novel approach for developing AE models that can integrate external knowledge sources into the learning process.<n>The proposed model is evaluated on three large-scale datasets from three scientific fields.
  arXiv  Detail & Related papers  (2023-03-12T18:00:12Z)
• Part-Based Models Improve Adversarial Robustness [57.699029966800644]
  We show that combining human prior knowledge with end-to-end learning can improve the robustness of deep neural networks. Our model combines a part segmentation model with a tiny classifier and is trained end-to-end to simultaneously segment objects into parts. Our experiments indicate that these models also reduce texture bias and yield better robustness against common corruptions and spurious correlations.
  arXiv  Detail & Related papers  (2022-09-15T15:41:47Z)

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.