Machine Learning Expedition

Deep Learning, Computer Vision, Machine Learning

[Machine Learning Expedition @MLexpAI]

Agents communicating through latent states instead of natural language
https://t.co/FGCWbV24Ra

[Machine Learning Expedition @MLexpAI]

Paper: https://t.co/a2nHLCl0kH

[Machine Learning Expedition @MLexpAI]

MEMRL: SELF-EVOLVING AGENTS VIA RUNTIME REINFORCEMENT LEARNING ON EPISODIC MEMORY

1. Introduces a two-phase retrieval system that first filters memory candidates by semantic relevance and then selects the final ones based on learned utility (Q-values), moving beyond purely semantic matching.

[Machine Learning Expedition @MLexpAI]

Explicitly separates stable LLM reasoning from a plastic, evolving memory, effectively addressing the stability-plasticity dilemma to allow for continuous runtime improvement and avoid catastrophic forgetting.

[Machine Learning Expedition @MLexpAI]

https://t.co/X4Otz7JrAR

[Machine Learning Expedition @MLexpAI]

3. Modular and Multi-Path-Aware Offline Benchmarking for Mobile GUI Agents
https://t.co/PD9sU43nN5

[Machine Learning Expedition @MLexpAI]

Recent LLM Agent Papers
1. MAC: A Multi-Agent Framework for Interactive User Clarification in Multi-turn Conversations

[Machine Learning Expedition @MLexpAI]

2. WEBOPERATOR: ACTION-AWARE TREE SEARCH FOR AUTONOMOUS AGENTS IN WEB ENVIRONMENT
https://t.co/9XBK299qxg

[Machine Learning Expedition @MLexpAI]

Paper URL: https://t.co/bfkHmtqAVk

[Machine Learning Expedition @MLexpAI]

ToolOrchestra: Elevating Intelligence via Efficient Model and Tool Orchestration

[Machine Learning Expedition @MLexpAI]

Paper Link: https://t.co/HrIZJF33f2

[Machine Learning Expedition @MLexpAI]

Towards a Science of Scaling Agent Systems-
This paper aims to move agentic AI systems (LLM-powered systems that reason, plan, and act) from heuristic practice toward a principled, quantitative science—especially for scaling those systems effectively

[Machine Learning Expedition @MLexpAI]

• Capability Saturation

Once a single agent achieves ~45% task performance baseline, additional agents often degrade results because coordination overhead dominates.

[Machine Learning Expedition @MLexpAI]

This paper got best paper award from Neurips 2025

[Machine Learning Expedition @MLexpAI]

Authors introduce infinity-chat dataset with 26k real world open-ended questions. The also provide human annotations including absolute rating and pairwise preferences.
https://t.co/zrDqHjbE8d