Jul 2026· Signal Processing and Communications Applications Conference· pp. 1-4· 0 citations· 12 references
Abstract
Label noise, which is frequently encountered in real-world data, is a critical problem that can directly degrade model performance. In this study, we systematically investigated the impact of label noise on intent classification by comparing the in-context learning (ICL) approach of large language models (LLMs) with fine-tuned transformer models. In the experiments conducted on the Banking77 dataset, we evaluated four LLMs and two transformer models under three different noise types and three different noise levels. We also tested the LLMs with different prompting strategies to examine the effect of taking precautions against potential noise on different LLMs. Our findings show that strong LLMs experience less than 2 percent loss in accuracy and F1 score even under the heaviest noise conditions, whereas in fine-tuned transformer models and relatively weaker LLMs, the drop can reach the 15-20 percent range.
By thoroughly unifying 36 English hate speech datasets spanning multiple labeling schemes, this work fine-tune a generalist LLM, based on Qwen3 (Qwen Team, 2025), specifically for hate speech mitigation, demonstrating not only state-of-the-art performance on in-domain benchmarks but also substantial improvements in cross-domain and cross-lingual generalization--areas where encoder-based specialist classifiers often struggle.
Lukas Edman, Daryna Dementieva, Alexander Fraser· 0 citations
Test set results show that Decoding-Enhanced Bert with Disentangled Attention (DeBERTa) achieves the highest macro F1 − Score of 85.48%, surpassing the previously top-ranked Multi-Task Learning (MTL) system, which attains a macro F1 of 83.07%.
Batyr Sharimbayev, S. Kadyrov· Journal of Advances in Infor...· 0 citations
Treating supervision format as a first-class hyperparameter for multi-task reasoning SFT in large language models—at least in this benchmark-and-model setting—rather than a mere rendering detail is supported.
Nhat Thanh Vu, M. Rashid, Fariza Sabrina· Electronics· 0 citations
The degradation rate across neural models, both sentence embeddings and decoder-only LLMs, is studied, and how consistent it is depends on the scale of the noise: under word-level noise, models with very different architectures decline along nearly the same curve, while under character-level noise they separate.
Despite their strong performance, large language models remain highly sensitive to prompt formulation. Prior work addresses this through refined data construction or through dedicated robustness objectives. We reproduce and compare these strategies under controlled conditions, and measure how effective they are in addressing models'prompt sensitivity. We find the current robustness fine-tuning methods improve over standard fine-tuning and in-context learning, but the best-to-worst prompt gap remains as high as 40-57% of performance. Moreover, the recent robustness-enhancing methods we test - CoIN for contrastive alignment and PPCL for consistency regularization - often fail to outperform the simplest data construction strategy: training on one template per batch. Our diagnostics explain these results. The auxiliary objectives move the quantity they penalize, but do not generalize beyond it. Additionally, data construction strategies differ due to the conflicting signs of per-template gradients on 57-64% of parameters. Thus, batches that mix formulations force the optimizer to reconcile competing updates instead of finding a shared, prompt-agnostic one.
A hybrid detection framework which combines semantically deep embeddings from the RoBERTa transformer with a set of carefully designed language statistics and linguistic statistics and shows excellent resistance to the surface-level adversarial paraphrasing strategy.
Anita Rani, Suman· International Journal of Sci...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
