Nov 2026· Proceedings of the 34th ACM International Conference on Multimedia· 0 citations· 16 references
Abstract
Multimodal Large Language Models (MLLMs) have advanced video temporal grounding (VTG) by formulating timestamp prediction as autoregressive token generation. However, this discrete formulation is susceptible to temporal hallucination. When visual evidence is insufficient, models default to linguistic priors, yielding plausible but ungrounded timestamps. While Contrastive Decoding (CD) offers a promising training-free strategy for mitigating such hallucination, it operates uniformly across the entire vocabulary. Since accurate temporal grounding hinges on probability distinctions among a small set of numerical tokens, this uniform operation dilutes fine-grained numerical distinctions, often causing format collapse or numerically invalid outputs. In this paper, we diagnose this limitation through the grouping property of Shannon entropy, revealing that the entropy contribution from non-numerical tokens dominates the total predictive entropy. This entropy imbalance pushes the task-relevant numerical uncertainty to the margin, limiting the effectiveness of full-vocabulary CD. Building on this analysis, we propose Subspace Entropic Contrastive Decoding (SECD), which dynamically isolates the numerical subspace and confines contrastive operations to tokens that determine temporal boundaries. Since the strength of linguistic priors varies across the layer hierarchy, a single-layer reference alone is insufficient. SECD further computes renormalized entropy of the numerical subspace at multiple intermediate layers. These entropy values serve as dynamic coefficients to aggregate layer-wise predictions into a unified contrastive reference. Extensive experiments across five widely used VTG benchmarks demonstrate that SECD, as a training-free and plug-and-play module, effectively mitigates temporal hallucination and yields precise and visually grounded timestamp predictions.
Supporting data, adapters, predictions and code for the article *Low-Cost LoRA Fine-Tuning of Small Language Models for Multi-Step Arithmetic Reasoning* by Jake O'Grady, Asena Isik Gürhan, Chee Fong Ting and Effirul Ramlan (University of Galway). We generated 20,000 GSM8K-derived arithmetic problems with step-by-step s...
O'Grady, Jake, Gürhan, Asena Isik, Chee, Fong Ting et al.· Zenodo (CERN European Organi...· 465 citations
The results are packaged in the Greenfield Startup Model (GSM), which explains the priority of startups to release the product as quickly as possible, and the need to shorten time-to-market, by speeding up the development through low-precision engineering activities.
Carmine Giardino, Nicolò Paternoster, M. Unterkalmsteiner et al.· IEEE Transactions on Softwar...· 178 citations· ⚡14
Software startup companies develop innovative, software-intensive products within limited timeframes and with few resources, searching for sustainable and scalable business models.
M. Unterkalmsteiner, P. Abrahamsson, Xiaofeng Wang et al.· e-Informatica Software Engin...· 157 citations· ⚡17
This study conducts a case survey study based on the secondary data of the major pivots happened in 49 software startups, and demonstrates that customer need pivot is the most common among all pivot types.
Sohaib Shahid Bajwa, Xiaofeng Wang, Anh Nguyen-Duc et al.· Empirical Software Engineeri...· 127 citations· ⚡15
The comparison of adopter and non-adopter sample reveals three potential adoption inhibitor, security, data privacy, and portability, which underlines the importance of the technical and security perspectives for research investigating the adoption of technology.
Nattakarn Phaphoom, Xiaofeng Wang, S. Samuel et al.· Journal of Systems and Softw...· 111 citations· ⚡8
Objective There is dearth of information on the timeliness of antenatal care (ANC) uptake. This study aimed to determine the timely ANC uptake by a medically trained provider (MTP) as per the World Health Organization (WHO) recommendations and the country guideline. Methods Cross-sectional survey was done with 2,731 wo...
B. Sarker, Musfikur Rahman, T. Rahman et al.· PLoS ONE· 89 citations· ⚡7
Related blog posts
Microsoft Research Blog· microsoft.comNov 16, 2026
AI coding agents are increasingly help developers fix bugs and build features largely on their own, but every interaction with their models consumes tokens. In each session, an agent sends instructions, tool descriptions, and session history to its model. Most efforts to reduce this cost focus on individual tasks. Yet developers return to the same […] The post TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient appeared first on Microsoft Research.
MIT scientists developed a training method to generate models that simulate the effects of human activities on climate much more quickly than other approaches.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.