8 papers
Direct Pass@ Optimization for Thought Pruning: Balancing Quality and Diversity in Test-Time Reasoning
Mahdi Salmani, Adel Javanmard, Devansh Gupta +1
Test-time scaling has emerged as a key inference strategy for reasoning models, where a model generates a large number of candidate solutions in the hope of finding at least one co…
A Gold-Standard Study of What Makes a Lightweight Game-Playing Agent Strong
Nima Kelidari, Mohammadsaeed Haghi, Mahdi Salmani
Reinforcement learning agents for imperfect-information card games are only as strong as the opponents they train against, and they are hard to grade, since they beat a random oppo…
Differentiating Through Dual Prices: End-to-End Policy Learning Under Capacity Constraints
Mohammadsaeed Haghi, Mahdi Salmani, Nima Kelidari +2
Many social services assign scarce resources, such as housing assistance or hospital interventions, to people who arrive one at a time. Each arrival must receive a decision immedia…
Early Stopping for Large Reasoning Models via Confidence Dynamics
Parsa Hosseini, Sumit Nawathe, Mahdi Salmani +2
Large reasoning models rely on long chain-of-thought generation to solve complex problems, but extended reasoning often incurs substantial computational cost and can even degrade p…
Sampling and Loss Weights in Multi-Domain Training
Mahdi Salmani, Pratik Worah, Meisam Razaviyayn +1
In the training of large deep neural networks, there is a need for vast amounts of training data. To meet this need, data is collected from multiple domains, such as Wikipedia and…
From Filters to VLMs: Benchmarking Defogging Methods through Object Detection and Segmentation Performance
Ardalan Aryashad, Parsa Razmara, Amin Mahjoub +3
Autonomous driving perception systems are particularly vulnerable in foggy conditions, where light scattering reduces contrast and obscures fine details critical for safe operation…
Rewriting the Budget: A General Framework for Black-Box Attacks Under Cost Asymmetry
Mahdi Salmani, Alireza Abdollahpoorrostam, Seyed-Mohsen Moosavi-Dezfooli
Traditional decision-based black-box adversarial attacks on image classifiers aim to generate adversarial examples by slightly modifying input images while keeping the number of qu…
Blacksmith: Fast Adversarial Training of Vision Transformers via a Mixture of Single-step and Multi-step Methods
Mahdi Salmani, Alireza Dehghanpour Farashah, Mohammad Azizmalayeri +4
Despite the remarkable success achieved by deep learning algorithms in various domains, such as computer vision, they remain vulnerable to adversarial perturbations. Adversarial Tr…