activity
20242026
collaborators

10 papers

cs.AI2026

Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity

Menglin Xia, Xuchao Zhang, Shantanu Dixit +6

Agent memory systems must accommodate continuously growing information while supporting efficient, context-aware retrieval for downstream tasks. Abstraction is essential for scalin…

cs.LG2026

Revisiting Transformer Layer Parameterization Through Causal Energy Minimization

Jin Xu, Camille Couturier, Victor Rühle +2

Transformer blocks typically combine multi-head attention (MHA) for token mixing with gated MLPs for token-wise feature transformation, yet many choices in their parameterization r…

cs.CL2025

Cost-Aware Retrieval-Augmentation Reasoning Models with Adaptive Retrieval Depth

Helia Hashemi, Victor Rühle, Saravan Rajmohan

Reasoning models have gained significant attention due to their strong performance, particularly when enhanced with retrieval augmentation. However, these models often incur high c…

cs.AI2025

LEGOMem: Modular Procedural Memory for Multi-agent LLM Systems for Workflow Automation

Dongge Han, Camille Couturier, Daniel Madrigal Diaz +3

We introduce LEGOMem, a modular procedural memory framework for multi-agent large language model (LLM) systems in workflow automation. LEGOMem decomposes past task trajectories int…

cs.CL2025

OdysseyBench: Evaluating LLM Agents on Long-Horizon Complex Office Application Workflows

Weixuan Wang, Dongge Han, Daniel Madrigal Diaz +3

Autonomous agents powered by large language models (LLMs) are increasingly deployed in real-world applications requiring complex, long-horizon workflows. However, existing benchmar…

cs.LG2025

BEST-Route: Adaptive LLM Routing with Test-Time Optimal Compute

Dujian Ding, Ankur Mallick, Shaokun Zhang +7

Large language models (LLMs) are powerful tools but are often expensive to deploy at scale. LLM query routing mitigates this by dynamically assigning queries to models of varying c…