Showing 2026Show all
2 papers · 1 filter
cs.LG2026
Cost-Aware Best-LLM Identification using Dueling Feedback
Sarvesh Gharat, Nikhil Karamchandani, Jayakrishnan Nair
Inspired by the problem of identifying the best model from a collection of large language models (LLMs) with heterogeneous querying costs, we formulate and analyse a variant of the…
cs.LG2026
Sequence prediction under a lying oracle
Puspabeethi Samanta, Nikhil Karamchandani, Jayakrishnan Nair
We consider the problem of sequential prediction of an -ary sequence, where at each epoch, (i) the environment selects an outcome from an -ary alphabet, (ii) the learner sele…