Showing cs.LGShow all
3 papers · 1 filter
cs.LG2026
SemanticZip: A Pilot Framework for Lossy Text Compression with LLMs as Semantic Decompressors
Natalia Trukhina, Vadim Vashkelis
Text compression for large language model (LLM) systems is usually framed as token deletion, retrieval, summarization, or exact reconstruction. We study a more aggressive but expli…
cs.LG2026
Compress the Context, Keep the Commitments: A Formal Framework for Verifiable LLM Context Compression
Natalia Trukhina, Vadim Vashkelis
LLM context is not just tokens; it is a set of commitments. Long-running conversations accumulate goals, constraints, decisions, preferences, tool results, retrieved evidence, arti…
cs.LG2026
HI-MoE: Hierarchical Instance-Conditioned Mixture-of-Experts for Object Detection
Vadim Vashkelis, Natalia Trukhina
Mixture-of-Experts (MoE) architectures enable conditional computation by activating only a subset of model parameters for each input. Although sparse routing has been highly effect…