Classifier Language Models: Unifying Sparse Finetuning and Adaptive Tokenization for Specialized Classification Tasks

• URL: http://arxiv.org/abs/2508.08635v1
• Date: Tue, 12 Aug 2025 04:59:01 GMT
• Title: Classifier Language Models: Unifying Sparse Finetuning and Adaptive Tokenization for Specialized Classification Tasks
• Authors: Adit Krishnan, Chu Wang, Chris Kong,
• Abstract summary: We develop a token-driven sparse finetuning strategy to adapt small language models to specialized classification tasks.<n>We identify and finetune a small sensitive subset of model parameters by leveraging task-specific token constructs in the finetuning dataset.<n>We achieve greater stability and half the training costs vs. end-to-end finetuning.
• Score: 5.857929080874287
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Semantic text classification requires the understanding of the contextual significance of specific tokens rather than surface-level patterns or keywords (as in rule-based or statistical text classification), making large language models (LLMs) well-suited for this task. However, semantic classification applications in industry, like customer intent detection or semantic role labeling, tend to be highly specialized. They require annotation by domain experts in contrast to general-purpose corpora for pretraining. Further, they typically require high inference throughputs which limits the model size from latency and cost perspectives. Thus, for a range of specialized classification tasks, the preferred solution is to develop customized classifiers by finetuning smaller language models (e.g., mini-encoders, small language models). In this work, we develop a token-driven sparse finetuning strategy to adapt small language models to specialized classification tasks. We identify and finetune a small sensitive subset of model parameters by leveraging task-specific token constructs in the finetuning dataset, while leaving most of the pretrained weights unchanged. Unlike adapter approaches such as low rank adaptation (LoRA), we do not introduce additional parameters to the model. Our approach identifies highly relevant semantic tokens (case study in the Appendix) and outperforms end-to-end finetuning, LoRA, layer selection, and prefix tuning on five diverse semantic classification tasks. We achieve greater stability and half the training costs vs. end-to-end finetuning.

Related papers

• GLiClass: Generalist Lightweight Model for Sequence Classification Tasks [49.2639069781367]
  We propose GLiClass, a novel method that adapts the GLiNER architecture for sequence classification tasks.<n>Our approach achieves strong accuracy and efficiency comparable to embedding-based methods, while maintaining the flexibility needed for zero-shot and few-shot learning scenarios.
  arXiv  Detail & Related papers  (2025-08-11T06:22:25Z)
• Large Language Models in the Task of Automatic Validation of Text Classifier Predictions [55.2480439325792]
  Machine learning models for text classification are trained to predict a class for a given text.<n>To do this, training and validation samples must be prepared, and each text is assigned a class.<n>Human annotators are usually assigned by human annotators with different expertise levels, depending on the specific classification task.<n>This paper proposes several approaches to replace human annotators with Large Language Models.
  arXiv  Detail & Related papers  (2025-05-24T13:19:03Z)
• Self-Regularization with Sparse Autoencoders for Controllable LLM-based Classification [29.74457390987092]
  We propose a novel framework to identify and regularize unintended features in large language models (LLMs) latent spaces.<n>We evaluate the proposed framework on three real-world tasks, including toxic chat detection, reward modeling, and disease diagnosis.
  arXiv  Detail & Related papers  (2025-02-19T22:27:59Z)
• SciPrompt: Knowledge-augmented Prompting for Fine-grained Categorization of Scientific Topics [2.3742710594744105]
  We introduce SciPrompt, a framework designed to automatically retrieve scientific topic-related terms for low-resource text classification tasks. Our method outperforms state-of-the-art, prompt-based fine-tuning methods on scientific text classification tasks under few and zero-shot settings.
  arXiv  Detail & Related papers  (2024-10-02T18:45:04Z)
• Language Models for Text Classification: Is In-Context Learning Enough? [54.869097980761595]
  Recent foundational language models have shown state-of-the-art performance in many NLP tasks in zero- and few-shot settings. An advantage of these models over more standard approaches is the ability to understand instructions written in natural language (prompts) This makes them suitable for addressing text classification problems for domains with limited amounts of annotated instances.
  arXiv  Detail & Related papers  (2024-03-26T12:47:39Z)
• Prompt Tuned Embedding Classification for Multi-Label Industry Sector Allocation [2.024620791810963]
  This study benchmarks the performance of Prompt Tuning and baselines for multi-label text classification. It is applied to classifying companies into an investment firm's proprietary industry taxonomy. We confirm that the model's performance is consistent across both well-known and less-known companies.
  arXiv  Detail & Related papers  (2023-09-21T13:45:32Z)
• Towards Realistic Zero-Shot Classification via Self Structural Semantic Alignment [53.2701026843921]
  Large-scale pre-trained Vision Language Models (VLMs) have proven effective for zero-shot classification. In this paper, we aim at a more challenging setting, Realistic Zero-Shot Classification, which assumes no annotation but instead a broad vocabulary. We propose the Self Structural Semantic Alignment (S3A) framework, which extracts structural semantic information from unlabeled data while simultaneously self-learning.
  arXiv  Detail & Related papers  (2023-08-24T17:56:46Z)
• Zero-Shot Text Classification via Self-Supervised Tuning [46.9902502503747]
  We propose a new paradigm based on self-supervised learning to solve zero-shot text classification tasks. tuning the language models with unlabeled data, called self-supervised tuning. Our model outperforms the state-of-the-art baselines on 7 out of 10 tasks.
  arXiv  Detail & Related papers  (2023-05-19T05:47:33Z)
• Attention is Not Always What You Need: Towards Efficient Classification of Domain-Specific Text [1.1508304497344637]
  For large-scale IT corpora with hundreds of classes organized in a hierarchy, the task of accurate classification of classes at the higher level in the hierarchies is crucial. In the business world, an efficient and explainable ML model is preferred over an expensive black-box model, especially if the performance increase is marginal. Despite the widespread use of PLMs, there is a lack of a clear and well-justified need to as why these models are being employed for domain-specific text classification.
  arXiv  Detail & Related papers  (2023-03-31T03:17:23Z)
• CCPrefix: Counterfactual Contrastive Prefix-Tuning for Many-Class Classification [57.62886091828512]
  We propose a brand-new prefix-tuning method, Counterfactual Contrastive Prefix-tuning (CCPrefix) for many-class classification. Basically, an instance-dependent soft prefix, derived from fact-counterfactual pairs in the label space, is leveraged to complement the language verbalizers in many-class classification.
  arXiv  Detail & Related papers  (2022-11-11T03:45:59Z)
• On Guiding Visual Attention with Language Specification [76.08326100891571]
  We use high-level language specification as advice for constraining the classification evidence to task-relevant features, instead of distractors. We show that supervising spatial attention in this way improves performance on classification tasks with biased and noisy data.
  arXiv  Detail & Related papers  (2022-02-17T22:40:19Z)
• Learning Debiased and Disentangled Representations for Semantic Segmentation [52.35766945827972]
  We propose a model-agnostic and training scheme for semantic segmentation. By randomly eliminating certain class information in each training iteration, we effectively reduce feature dependencies among classes. Models trained with our approach demonstrate strong results on multiple semantic segmentation benchmarks.
  arXiv  Detail & Related papers  (2021-10-31T16:15:09Z)

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.