Functional-level Uncertainty Quantification for Calibrated Fine-tuning on LLMs

• URL: http://arxiv.org/abs/2410.06431v3
• Date: Sun, 25 May 2025 08:11:40 GMT
• Title: Functional-level Uncertainty Quantification for Calibrated Fine-tuning on LLMs
• Authors: Ruijia Niu, Dongxia Wu, Rose Yu, Yi-An Ma,
• Abstract summary: Calibrated Fine-Tuning (UQ4CT) captures and calibrates uncertainty over the space of functions that map input prompts to outputs.<n>We implement UQ4CT during the fine-tuning stage via a mixture-of-experts framework that hierarchically decomposes the functional space.<n>Even under distribution shift, UQ4CT maintains superior ECE performance with high accuracy, showcasing improved generalizability.
• Score: 21.94487480599671
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Accurate uncertainty quantification in large language models (LLMs) is essential for providing credible confidence estimates over their outputs. However, fine-tuned LLMs often exhibit overconfidence in uncertain predictions, which stems from their limited ability to generalize with sparse data. Existing parameter efficient fine-tuning (PEFT) uncertainty quantification methods for LLMs focus on post fine-tuning stage, and thus fail to address the core issue: limited specialization of PEFT adapters to accurately capture task-specific input-output relationships. To address these limitations, we propose Functional-Level Uncertainty Quantification for Calibrated Fine-Tuning (UQ4CT), which captures and calibrates uncertainty over the space of functions that map input prompts to outputs. We implement UQ4CT during the fine-tuning stage via a mixture-of-experts framework that hierarchically decomposes the functional space. Empirically, UQ4CT achieves over $25\%$ reduction in Expected Calibration Error (ECE) while preserving high accuracy across five benchmarks. Even under distribution shift, UQ4CT maintains superior ECE performance with high accuracy, showcasing improved generalizability.

Related papers

• COIN: Uncertainty-Guarding Selective Question Answering for Foundation Models with Provable Risk Guarantees [51.5976496056012]
  COIN is an uncertainty-guarding selection framework that calibrates statistically valid thresholds to filter a single generated answer per question.<n>COIN estimates the empirical error rate on a calibration set and applies confidence interval methods to establish a high-probability upper bound on the true error rate.<n>We demonstrate COIN's robustness in risk control, strong test-time power in retaining admissible answers, and predictive efficiency under limited calibration data.
  arXiv  Detail & Related papers  (2025-06-25T07:04:49Z)
• Supervised Optimism Correction: Be Confident When LLMs Are Sure [91.7459076316849]
  We establish a novel theoretical connection between supervised fine-tuning and offline reinforcement learning. We show that the widely used beam search method suffers from unacceptable over-optimism. We propose Supervised Optimism Correction, which introduces a simple yet effective auxiliary loss for token-level $Q$-value estimations.
  arXiv  Detail & Related papers  (2025-04-10T07:50:03Z)
• Quantize What Counts: Bit Allocation Insights Informed by Spectral Gaps in Keys and Values [57.54443445583921]
  We provide two novel theorems aimed at enhancing KV quantization methods.<n>Our first theorem, termed Key-Value Norm Disparity, states that the key weight matrices by nature carry richer information.<n>Our second theorem, Key-Driven Quantization, posits that prioritizing the quantization precision of keys over values induces significant improvements to the overall quantization performance.
  arXiv  Detail & Related papers  (2025-02-20T22:24:27Z)
• COPU: Conformal Prediction for Uncertainty Quantification in Natural Language Generation [14.461333001997449]
  Uncertainty Quantification (UQ) for Natural Language Generation (NLG) is crucial for assessing the performance of Large Language Models (LLMs) We propose ourmethod, a method that explicitly adds the ground truth to the candidate outputs and uses logit scores to measure nonconformity.
  arXiv  Detail & Related papers  (2025-02-18T07:25:12Z)
• Monty Hall and Optimized Conformal Prediction to Improve Decision-Making with LLMs [7.843594672029363]
  Con conformal prediction (CP) is a model-agnostic framework for distribution-free uncertainty quantification. We introduce CP-OPT, an optimization framework to learn scores that minimize set sizes while maintaining coverage. We also propose emphconformal revision of questions (CROQ) to revise the problem by narrowing down the available choices to those in the prediction set.
  arXiv  Detail & Related papers  (2024-12-31T17:33:12Z)
• GAQAT: gradient-adaptive quantization-aware training for domain generalization [54.31450550793485]
  We propose a novel Gradient-Adaptive Quantization-Aware Training (GAQAT) framework for DG. Our approach begins by identifying the scale-gradient conflict problem in low-precision quantization. Extensive experiments validate the effectiveness of the proposed GAQAT framework.
  arXiv  Detail & Related papers  (2024-12-07T06:07:21Z)
• Calibrating Deep Neural Network using Euclidean Distance [5.675312975435121]
  In machine learning, Focal Loss is commonly used to reduce misclassification rates by emphasizing hard-to-classify samples. High calibration error indicates a misalignment between predicted probabilities and actual outcomes, affecting model reliability. This research introduces a novel loss function called Focal Loss (FCL), designed to improve probability calibration while retaining the advantages of Focal Loss in handling difficult samples.
  arXiv  Detail & Related papers  (2024-10-23T23:06:50Z)
• Feature Clipping for Uncertainty Calibration [24.465567005078135]
  Modern deep neural networks (DNNs) often suffer from overconfidence, leading to miscalibration. We propose a novel post-hoc calibration method called feature clipping (FC) to address this issue. FC involves clipping feature values to a specified threshold, effectively increasing entropy in high calibration error samples.
  arXiv  Detail & Related papers  (2024-10-16T06:44:35Z)
• Calibrating Language Models with Adaptive Temperature Scaling [58.056023173579625]
  We introduce Adaptive Temperature Scaling (ATS), a post-hoc calibration method that predicts a temperature scaling parameter for each token prediction. ATS improves calibration by over 10-50% across three downstream natural language evaluation benchmarks compared to prior calibration methods.
  arXiv  Detail & Related papers  (2024-09-29T22:54:31Z)
• ConU: Conformal Uncertainty in Large Language Models with Correctness Coverage Guarantees [68.33498595506941]
  We introduce a novel uncertainty measure based on self-consistency theory. We then develop a conformal uncertainty criterion by integrating the uncertainty condition aligned with correctness into the CP algorithm. Empirical evaluations indicate that our uncertainty measure outperforms prior state-of-the-art methods.
  arXiv  Detail & Related papers  (2024-06-29T17:33:07Z)
• Low-rank finetuning for LLMs: A fairness perspective [54.13240282850982]
  Low-rank approximation techniques have become the de facto standard for fine-tuning Large Language Models. This paper investigates the effectiveness of these methods in capturing the shift of fine-tuning datasets from the initial pre-trained data distribution. We show that low-rank fine-tuning inadvertently preserves undesirable biases and toxic behaviors.
  arXiv  Detail & Related papers  (2024-05-28T20:43:53Z)
• Confidence-aware Fine-tuning of Sequential Recommendation Systems via Conformal Prediction [46.76846936581471]
  In Sequential Recommendation Systems (SRecsys), traditional training approaches that rely on Cross-Entropy (CE) loss often prioritize accuracy but fail to align well with user satisfaction metrics.<n>We propose textbfCPFT, a novel fine-tuning framework that integrates Conformal Prediction (CP)-based losses with CE loss to optimize accuracy alongside confidence that better aligns with widely used top-$K$ metrics.
  arXiv  Detail & Related papers  (2024-02-14T06:43:02Z)
• L4Q: Parameter Efficient Quantization-Aware Fine-Tuning on Large Language Models [5.304907804008533]
  We propose L4Q, a method that integrates Quantization-Aware Training (QAT) with Low-Rank Adaptation (LoRA) By employing a memory-optimized layer design, L4Q significantly reduces QAT's memory overhead, making its training cost comparable to LoRA. Our experiments demonstrate that this combined approach to quantization and fine-tuning achieves superior accuracy.
  arXiv  Detail & Related papers  (2024-02-07T14:35:05Z)
• Sparse is Enough in Fine-tuning Pre-trained Large Language Models [98.46493578509039]
  We propose a gradient-based sparse fine-tuning algorithm, named Sparse Increment Fine-Tuning (SIFT) We validate its effectiveness on a range of tasks including the GLUE Benchmark and Instruction-tuning.
  arXiv  Detail & Related papers  (2023-12-19T06:06:30Z)
• Memory-Efficient Fine-Tuning of Compressed Large Language Models via sub-4-bit Integer Quantization [27.79783067245817]
  Large language models (LLMs) face the challenges in fine-tuning and deployment due to their high memory demands and computational costs. This paper presents Efficient Adaptation and Quantization-aware (PEQA) - a simple yet effective method that combines the advantages of PEFT with quantized LLMs.
  arXiv  Detail & Related papers  (2023-05-23T15:20:01Z)
• Towards Accurate Post-Training Quantization for Vision Transformer [48.779346466374406]
  Existing post-training quantization methods still cause severe performance drops. APQ-ViT surpasses the existing post-training quantization methods by convincing margins.
  arXiv  Detail & Related papers  (2023-03-25T03:05:26Z)
• Few-Shot Calibration of Set Predictors via Meta-Learned Cross-Validation-Based Conformal Prediction [33.33774397643919]
  This paper introduces a novel meta-learning solution that aims at reducing the set prediction size. It builds on cross-validation-based CP, rather than the less efficient validation-based CP. It preserves formal per-task calibration guarantees, rather than less stringent task-marginal guarantees.
  arXiv  Detail & Related papers  (2022-10-06T17:21:03Z)
• Optimal Clipping and Magnitude-aware Differentiation for Improved Quantization-aware Training [8.106641866299377]
  Current practices rely on scalars to set clipping threshold scalars and cannot be shown to be optimal. We propose Optimally Clippeds And Vectors ( OCTAV), a algorithm to determine MSE-optimal clipping scalars. OCTAV finds optimal clipping scalars on the fly, for every tensor, at every iteration of the quantization-aware training (QAT) routine.
  arXiv  Detail & Related papers  (2022-06-13T22:15:21Z)
• Parameterized Temperature Scaling for Boosting the Expressive Power in Post-Hoc Uncertainty Calibration [57.568461777747515]
  We introduce a novel calibration method, Parametrized Temperature Scaling (PTS) We demonstrate that the performance of accuracy-preserving state-of-the-art post-hoc calibrators is limited by their intrinsic expressive power. We show with extensive experiments that our novel accuracy-preserving approach consistently outperforms existing algorithms across a large number of model architectures, datasets and metrics.
  arXiv  Detail & Related papers  (2021-02-24T10:18:30Z)
• Amortized Conditional Normalized Maximum Likelihood: Reliable Out of Distribution Uncertainty Estimation [99.92568326314667]
  We propose the amortized conditional normalized maximum likelihood (ACNML) method as a scalable general-purpose approach for uncertainty estimation. Our algorithm builds on the conditional normalized maximum likelihood (CNML) coding scheme, which has minimax optimal properties according to the minimum description length principle. We demonstrate that ACNML compares favorably to a number of prior techniques for uncertainty estimation in terms of calibration on out-of-distribution inputs.
  arXiv  Detail & Related papers  (2020-11-05T08:04:34Z)

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.