Papers with model calibration

20 papers
Page Stream Segmentation with LLMs: Challenges and Applications in Insurance Document Automation (2025.coling-industry)

Copied to clipboard

Challenge: Existing models for page Stream Segmentation (PSS) are inadequate for high-stakes applications.
Approach: They propose to use large language models for page Stream Segmentation to apply parameter-efficient fine-tuning to real-world insurance data.
Outcome: The proposed model outperforms baseline models in page- and stream-level segmentation accuracy.
A Close Look into the Calibration of Pre-trained Language Models (2023.acl-long)

Copied to clipboard

Challenge: Pre-trained language models (PLMs) may fail in giving reliable estimates of their predictive uncertainty.
Approach: They conduct fine-grained control experiments to study the dynamic change in PLMs’ calibration performance in training.
Outcome: The proposed methods significantly reduce PLMs’ confidence in wrong predictions.
CATfOOD: Counterfactual Augmented Training for Improving Out-of-Domain Performance and Calibration (2024.eacl-long)

Copied to clipboard

Challenge: Large language models (LLMs) have shown remarkable generalization capabilities, performing well on various tasks such as question answering (QA), complex reasoning, and code generation.
Approach: They propose to augment training data of smaller language models with automatically generated counterfactuals (CF) instances to improve out-of-domain (OOD) performance of SLMs in extractive question answering setup.
Outcome: The proposed approach improves out-of-domain (OOD) performance of small language models in extractive question answering setup.
Extracting Victim Counts from Text (2023.eacl-main)

Copied to clipboard

Challenge: Using tagging and regex methods, data on injured, displaced, or abused victims is difficult . data on earthquake injuries and deaths is scarce, subjective, or biased .
Approach: They compare tagging approaches to extract injured, displaced, or abused victims . they discuss calibration and investigate out-of-distribution and few-shot performance .
Outcome: The proposed model is among the first to apply numeracy-focused large language models in a real-world use case with a positive impact.
Calibrating Zero-shot Cross-lingual (Un-)structured Predictions (2022.emnlp-main)

Copied to clipboard

Challenge: Existing need for model calibration when natural language models are deployed in critical tasks.
Approach: They compare model calibration methods in a context of zero-shot cross-lingual transfer with pre-trained language models.
Outcome: The proposed method fails to calibrate more complex confidence estimations in structured predictions compared to expressive alternatives like Gaussian Process Calibration.
Reasoning about Uncertainty: Do Reasoning Models Know When They Don’t Know? (2026.findings-eacl)

Copied to clipboard

Challenge: Reasoning models are prone to generating confident, plausible responses that are incorrect (hallucinations).
Approach: They introduce introspective uncertainty quantification to examine whether reasoning models are well-calibrated and does deeper reasoning improve their calibration?
Outcome: The proposed model calibrations show that models are overconfident, overconfent and overconfust with deeper reasoning.
STAR: Constraint LoRA with Dynamic Active Learning for Data-Efficient Fine-Tuning of Large Language Models (2024.findings-acl)

Copied to clipboard

Challenge: Existing studies show that supervised training is still necessary for complex reasoning tasks.
Approach: They propose a method to integrate uncertainty-based active learning and LoRA to effectively integrate the two methods.
Outcome: The proposed approach outperforms baseline models on three reasoning tasks.
RoAST: Robustifying Language Models via Adversarial Perturbation with Selective Training (2023.findings-emnlp)

Copied to clipboard

Challenge: Several perspectives of robustness for pre-trained language models have been studied independently, but lacking a unified consideration in multiple perspectives.
Approach: They propose a technique to enhance the multi-perspective robustness of LMs by introducing adversarial perturbation while the model parameters are selectively updated upon their relative importance.
Outcome: The proposed technique improves the robustness of LMs by incorporating four perspectives on model robustness.
Test-time Augmentation for Factual Probing (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing methods to improve factual probing are relation-specific and do not generalize to unseen relation types.
Approach: They propose to use test-time augmentation to augment and ensemble prompts at test time to reduce sensitivity to prompt variations.
Outcome: The proposed method improves model confidence, but for other models, it leads to degradation.
On the Calibration of Pre-trained Language Models using Mixup Guided by Area Under the Margin and Saliency (2022.acl-long)

Copied to clipboard

Challenge: Existing studies have shown that mixing up can improve model calibration on image classification tasks, but little is known about using it on natural language understanding (NLU) tasks.
Approach: They propose a mixup strategy for pre-trained language models that improves model calibration further by using the AUM statistic and saliency map.
Outcome: The proposed mixup improves model calibration on natural language understanding tasks while maintaining competitive accuracy.
On the Risk of Evidence Pollution for Malicious Social Text Detection in the Era of LLMs (2025.acl-long)

Copied to clipboard

Challenge: Evidence-enhanced detectors are able to detect malicious social text, but they are prone to evidence pollution.
Approach: They propose three defense strategies to mitigate evidence pollution by large language models by machine-generated text detection and a mixture of experts.
Outcome: The proposed defense strategies could mitigate evidence pollution, but they faced limitations for practical employment.
Boundary Smoothing for Named Entity Recognition (2022.acl-long)

Copied to clipboard

Challenge: Named entity recognition models often encounter over-confidence issues . boundary smoothing is a method that re-assigns entity probabilities from annotated spans to the surrounding ones .
Approach: They propose a method for regularizing entity probabilities from annotated spans to the surrounding ones.
Outcome: The proposed method achieves better than or competitive with previous state-of-the-art systems on well-known benchmarks.
Beyond Noise: Mitigating the Impact of Fine-grained Semantic Divergences on Neural Machine Translation (2021.acl-long)

Copied to clipboard

Challenge: Prior work treats all types of mismatches between source and target as noise . Consequently, it remains unclear how noisy parallel training samples impact NMT training.
Approach: They propose a divergent-aware NMT framework that uses factors to help NMT recover from the degradation caused by naturally occurring divergences.
Outcome: The proposed framework improves translation quality and model calibration on EN-FR tasks.
Local Temperature Beam Search: Avoid Neural Text DeGeneration via Enhanced Calibration (2023.findings-acl)

Copied to clipboard

Challenge: Existing approaches to inference have been based on stochastic decoding but they sacrifice output quality due to randomness.
Approach: They propose a deterministic decoding scheme, local temperature beam search, which reduces repetition while maintaining the level of coherence as in beam search.
Outcome: The proposed inference scheme reduces repetition while maintaining coherence as in beam search.
On the Calibration of Large Language Models and Alignment (2023.findings-emnlp)

Copied to clipboard

Challenge: Large language models are becoming more popular and are proving to be reliable . however, their reliability is often understudied due to their uncertainty and complex structure .
Approach: They conduct a systematic examination of the calibration of aligned language models throughout the entire construction process including pretraining and alignment training.
Outcome: The results shed light on whether popular large language models are well-calibrated and how the training process influences model calibration.
Hard Gate Knowledge Distillation - Leverage Calibration for Robust and Reliable Language Model (2022.emnlp-main)

Copied to clipboard

Challenge: Existing knowledge distillation schemes focus on a teacher as a source of knowledge and a gauge to detect miscalibration of a student.
Approach: They propose a method that uses a teacher model as a source of knowledge and a model as an error detector to detect miscalibration of a student.
Outcome: The proposed scheme improves model generalization and significantly lowers calibration error.
GRACE: A Granular Benchmark for Evaluating Model Calibration against Human Calibration (2025.acl-long)

Copied to clipboard

Challenge: Language models are often miscalibrated, leading to confidently incorrect answers.
Approach: They propose a benchmark for language model calibration that incorporates comparison with human calibration.
Outcome: The proposed metric analyzes model calibration errors and identifies types of miscalibration that differ from human behavior.
Mind the Gap: A Closer Look at Tokenization for Multiple-Choice Question Answering with LLMs (2025.emnlp-main)

Copied to clipboard

Challenge: Recent studies have highlighted the significant performance variation that can arise from minor changes in prompt design.
Approach: They propose to tokenize the space following the colon to facilitate automated answer extraction via next-token probabilities.
Outcome: The proposed tokenization improves model calibration and improves confidence estimates.
COM-BOM: Bayesian Exemplar Search for Efficiently Exploring the Accuracy-Calibration Pareto Frontier (2025.emnlp-main)

Copied to clipboard

Challenge: Prior exemplar selection methods focus on maximizing predictive accuracy, neglecting model calibration.
Approach: They propose to use a Bayesian optimization algorithm to optimize for predictive accuracy and calibration.
Outcome: The proposed algorithm beats or matches baselines on multiple tasks from un-saturated MMLU-pro benchmarks while requiring minimal number of API calls.
Model Calibration for Emotion Detection (2025.findings-emnlp)

Copied to clipboard

Challenge: a MixUp method is used to calibrate emotion detection models based on knowledge distillation and the MixUp data augmentation technique.
Approach: They propose a method that uses knowledge distillation and the MixUp data augmentation technique to enhance the trustworthiness of emotion detection models.
Outcome: The proposed method improves the accuracy of the teacher models and the student models.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations