5 papers
Knowing When to Ask: Segment-Level Credit Assignment for LLM Tool Use
Abhijit Kumar, Zoey Wu, Mohit Suley
Humans know when to reach for help e.g. warrants a calculator while does not. Language models do not. Prompt-based approaches can instruct a model when to inv…
Execution-Grounded Credit Assignment for GRPO in Code Generation
Abhijit Kumar, Natalya Kumar, Shikhar Gupta
Critic-free reinforcement learning with verifiable rewards (RLVR) improves code generation by optimizing unit-test pass rates, but GRPO-style updates suffer from coarse credit assi…
Hierarchical Resolution Transformers: A Wavelet-Inspired Architecture for Multi-Scale Language Understanding
Ayan Sar, Sampurna Roy, Kanav Gupta +3
Transformer architectures have achieved state-of-the-art performance across natural language tasks, yet they fundamentally misrepresent the hierarchical nature of human language by…
Dynamic Reasoning Chains through Depth-Specialized Mixture-of-Experts in Transformer Architectures
Sampurna Roy, Ayan Sar, Anurag Kaushish +3
Contemporary transformer architectures apply identical processing depth to all inputs, creating inefficiencies and limiting reasoning quality. Simple factual queries are subjected…
SwasthLLM: a Unified Cross-Lingual, Multi-Task, and Meta-Learning Zero-Shot Framework for Medical Diagnosis Using Contrastive Representations
Ayan Sar, Pranav Singh Puri, Sumit Aich +2
In multilingual healthcare environments, automatic disease diagnosis from clinical text remains a challenging task due to the scarcity of annotated medical data in low-resource lan…