activity
20242026
collaborators

7 papers

cs.LG2026

From Markov to Laplace: How Mamba In-Context Learns Markov Chains

Marco Bondaschi, Nived Rajaraman, Xiuying Wei +5

While transformer-based language models have driven the AI revolution thus far, their computational complexity has spurred growing interest in viable alternatives, such as structur…

cs.LG2025

What One Cannot, Two Can: Two-Layer Transformers Provably Represent Induction Heads on Any-Order Markov Chains

Chanakya Ekbote, Marco Bondaschi, Nived Rajaraman +4

In-context learning (ICL) is a hallmark capability of transformers, through which trained models learn to adapt to new tasks by leveraging information from the input context. Prior…

cs.IT2025

The Conditional Regret-Capacity Theorem for Batch Universal Prediction

Marco Bondaschi, Michael Gastpar

We derive a conditional version of the classical regret-capacity theorem. This result can be used in universal prediction to find lower bounds on the minimal batch regret, which is…

cs.LG2025

Attention with Markov: A Framework for Principled Analysis of Transformers via Markov Chains

Ashok Vardhan Makkuva, Marco Bondaschi, Adway Girish +4

Attention-based transformers have achieved tremendous success across a variety of disciplines including natural languages. To deepen our understanding of their sequential modeling…

cs.IT2024

Alpha-NML Universal Predictors

Marco Bondaschi, Michael Gastpar

Inspired by the connection between classical regret measures employed in universal prediction and Rényi divergence, we introduce a new class of universal predictors that depend on…

cs.LG2024

Fundamental Limits of Prompt Compression: A Rate-Distortion Framework for Black-Box Language Models

Alliot Nagle, Adway Girish, Marco Bondaschi +3

We formalize the problem of prompt compression for large language models (LLMs) and present a framework to unify token-level prompt compression methods which create hard prompts fo…