4 papers
Closing the Calibration Gap in Semantic Caching
Aditeya Baral, Radoslav Ralev, Iliya Sotirov Zhechev +2
Semantic caching cuts LLM inference costs by serving a cached response to semantically similar queries. Standard practice evaluates these systems using PR-AUC, a metric that only m…
Training for Compositional Sensitivity Reduces Dense Retrieval Generalization
Radoslav Ralev, Aditeya Baral, Iliya Zhechev +2
Dense retrieval compresses texts into single embeddings ranked by cosine similarity. While efficient for recall, this interface is brittle for identity-level matching: minimal comp…
Can LLMs Math? -- Exploring the Pitfalls in Mathematical Reasoning
Tiasa Singha Roy, Aditeya Baral, Ayush Rajesh Jhaveri +1
Large language models (LLMs) demonstrate considerable potential in various natural language tasks but face significant challenges in mathematical reasoning, particularly in executi…
CMLFormer: A Dual Decoder Transformer with Switching Point Learning for Code-Mixed Language Modeling
Aditeya Baral, Allen George Ajith, Roshan Nayak +1
Code-mixed languages, characterized by frequent within-sentence language transitions, present structural challenges that standard language models fail to address. In this work, we…