Tootfinder

A break from the norm? Parametric representations of preference heterogeneity for discrete choice models in health
John Buckell, Alice Wreford, Matthew Quaife, Thomas O. Hancock
https://arxiv.org/abs/2506.14099

A break from the norm? Parametric representations of preference heterogeneity for discrete choice models in health
Background: Any sample of individuals has its own, unique distribution of preferences for choices that they make. Discrete choice models try to capture these distributions. Mixed logits are by far the most commonly used choice model in health. A raft of parametric model specifications for these models are available. We test a range of alternatives assumptions, and model averaging, to test if or how model outputs are impacted. Design: Scoping review of current modelling practices. Seven alternat…

Jointly Generating and Attributing Answers using Logits of Document-Identifier Tokens
Lucas Albarede, Jose Moreno, Lynda Tamine, Luce Lefeuvre
https://arxiv.org/abs/2508.08942 h…

Jointly Generating and Attributing Answers using Logits of Document-Identifier Tokens
Despite their impressive performances, Large Language Models (LLMs) remain prone to hallucination, which critically undermines their trustworthiness. While most of the previous work focused on tackling answer and attribution correctness, a recent line of work investigated faithfulness, with a focus on leveraging internal model signals to reflect a model's actual decision-making process while generating the answer. Nevertheless, these methods induce additional latency and have shown limitations …

A Stable and Principled Loss Function for Direct Language Model Alignment
Yuandong Tan
https://arxiv.org/abs/2508.07137 https://arxiv.org/pdf/2508.07137

A Stable and Principled Loss Function for Direct Language Model Alignment
The alignment of large language models (LLMs) with human preferences is commonly achieved through Reinforcement Learning from Human Feedback (RLHF). Direct Preference Optimization (DPO) simplified this paradigm by establishing a direct mapping between the optimal policy and a reward function, eliminating the need for an explicit reward model. However, we argue that the DPO loss function is theoretically misaligned with its own derivation, as it promotes the indefinite maximization of a logits d…

Mitigating Hallucination of Large Vision-Language Models via Dynamic Logits Calibration
Jiahe Chen, Jiaying He, Qian Shao, Qiyuan Chen, Jiahe Ying, Hongxia Xu, Jintai Chen, Jianwei Zheng, Jian Wu
https://arxiv.org/abs/2506.21509

Mitigating Hallucination of Large Vision-Language Models via Dynamic Logits Calibration
Large Vision-Language Models (LVLMs) have demonstrated significant advancements in multimodal understanding, yet they are frequently hampered by hallucination-the generation of text that contradicts visual input. Existing training-free decoding strategies exhibit critical limitations, including the use of static constraints that do not adapt to semantic drift during generation, inefficiency stemming from the need for multiple forward passes, and degradation of detail due to overly rigid interve…

DeCRED: Decoder-Centric Regularization for Encoder-Decoder Based Speech Recognition
Alexander Polok, Santosh Kesiraju, Karel Bene\v{s}, Bolaji Yusuf, Luk\'a\v{s} Burget, Jan \v{C}ernock\'y
https://arxiv.org/abs/2508.08938

DeCRED: Decoder-Centric Regularization for Encoder-Decoder Based Speech Recognition
This paper presents a simple yet effective regularization for the internal language model induced by the decoder in encoder-decoder ASR models, thereby improving robustness and generalization in both in- and out-of-domain settings. The proposed method, Decoder-Centric Regularization in Encoder-Decoder (DeCRED), adds auxiliary classifiers to the decoder, enabling next token prediction via intermediate logits. Empirically, DeCRED reduces the mean internal LM BPE perplexity by 36.6% relative to 11…

Joint Feature and Output Distillation for Low-complexity Acoustic Scene Classification
Haowen Li, Ziyi Yang, Mou Wang, Ee-Leng Tan, Junwei Yeow, Santi Peksi, Woon-Seng Gan
https://arxiv.org/abs/2507.19557

Joint Feature and Output Distillation for Low-complexity Acoustic Scene Classification
This report presents a dual-level knowledge distillation framework with multi-teacher guidance for low-complexity acoustic scene classification (ASC) in DCASE2025 Task 1. We propose a distillation strategy that jointly transfers both soft logits and intermediate feature representations. Specifically, we pre-trained PaSST and CP-ResNet models as teacher models. Logits from teachers are averaged to generate soft targets, while one CP-ResNet is selected for feature-level distillation. This enables…

Bag of Coins: A Statistical Probe into Neural Confidence Structures
Agnideep Aich, Ashit Baran Aich, Md Monzur Murshed, Sameera Hewage, Bruce Wade
https://arxiv.org/abs/2507.19774

Bag of Coins: A Statistical Probe into Neural Confidence Structures
Modern neural networks, despite their high accuracy, often produce poorly calibrated confidence scores, limiting their reliability in high-stakes applications. Existing calibration methods typically post-process model outputs without interrogating the internal consistency of the predictions themselves. In this work, we introduce a novel, non-parametric statistical probe, the Bag-of-Coins (BoC) test, that examines the internal consistency of a classifier's logits. The BoC test reframes confidenc…

Analyzing and Mitigating Repetitions in Trip Recommendation
Wenzheng Shu, Kangqi Xu, Wenxin Tai, Ting Zhong, Yong Wang, Fan Zhou
https://arxiv.org/abs/2507.19798 https://…

Analyzing and Mitigating Repetitions in Trip Recommendation
Trip recommendation has emerged as a highly sought-after service over the past decade. Although current studies significantly understand human intention consistency, they struggle with undesired repetitive outcomes that need resolution. We make two pivotal discoveries using statistical analyses and experimental designs: (1) The occurrence of repetitions is intricately linked to the models and decoding strategies. (2) During training and decoding, adding perturbations to logits can reduce repeti…

Reducing Object Hallucination in Large Audio-Language Models via Audio-Aware Decoding
Tzu-wen Hsu, Ke-Han Lu, Cheng-Han Chiang, Hung-yi Lee
https://arxiv.org/abs/2506.07233

Reducing Object Hallucination in Large Audio-Language Models via Audio-Aware Decoding
Large Audio-Language Models (LALMs) can take audio and text as the inputs and answer questions about the audio. While prior LALMs have shown strong performance on standard benchmarks, there has been alarming evidence that LALMs can hallucinate what is presented in the audio. To mitigate the hallucination of LALMs, we introduce Audio-Aware Decoding (AAD), a lightweight inference-time strategy that uses contrastive decoding to compare the token prediction logits with and without the audio context…

Efficient Interleaved Speech Modeling through Knowledge Distillation
Mohammadmahdi Nouriborji, Morteza Rohanian
https://arxiv.org/abs/2506.23670 https://…

Efficient Interleaved Speech Modeling through Knowledge Distillation
Current speech language models exceed the size and latency constraints of many deployment environments. We build compact, expressive speech generation models through layer-aligned distillation, matching hidden states, attention maps, and softened logits to compress large multimodal transformers by 3x with minimal loss in performance. We introduce TinyWave, a family of 2B-parameter models for speech-to-speech and interleaved speech-text generation, trained on 50,000 hours of public audio. TinyWa…

AeroGPT: Leveraging Large-Scale Audio Model for Aero-Engine Bearing Fault Diagnosis
Jiale Liu, Dandan Peng, Huan Wang, Chenyu Liu, Yan-Fu Li, Min Xie
https://arxiv.org/abs/2506.16225

AeroGPT: Leveraging Large-Scale Audio Model for Aero-Engine Bearing Fault Diagnosis
Aerospace engines, as critical components in aviation and aerospace industries, require continuous and accurate fault diagnosis to ensure operational safety and prevent catastrophic failures. While deep learning techniques have been extensively studied in this context, they output logits or confidence scores, necessitating post-processing to derive actionable insights. Furthermore, the potential of large-scale audio models in this domain remains largely untapped. To address these limitations, t…