Agents

LLM agents and the systems around them: harnesses and orchestration, tool use and MCP, multi-agent systems, agent memory and planning

600 articles · page 10 of 13

Agents

Scientific agents gain auditability through enforced, re-verifiable research workflows

The IdeaHorizon Team describes AfS, a platform for scientific projects spanning dozens of agent runs with only occasional human involvement. Rather than relying on model instructions alone, AfS mechanically enforces commitments, evidence handling, review boundaries, and memory rules. The paper illustrates these mechanisms through failure traces and two completed campaigns, but does not provide a comparative benchmark.

·2 min
Agents

Preference learning improves collaboration among teams of language-model agents

Liu et al. formulate preference learning for multi-agent language-model systems and introduce MAPL, which iteratively improves agents by comparing their joint solutions with alternatives from comparator models. Across writing, coding, tool-use and travel-planning tasks, MAPL improved collaboration quality and efficiency, with its learned-reward variant generally outperforming direct preference optimization.

·2 min
Agents

Multi-agent LLM scaling benefits depend on task type and aggregation method

Fortuna and Bertalanič introduce Steiner's taxonomy of group tasks to analyze multi-agent LLM scaling, modeling independently sampled agents as conditionally independent given an item. Testing 13 open-weight models on teams of up to 30 agents across disjunctive and compensatory benchmarks, they find that on disjunctive tasks, plurality voting captures almost none of the potential from larger teams (model predicts within 0.5 points of true accuracy), while multi-round revision yields strong gains that saturate with just one peer. On compensatory Fermi estimation, averaging reduces error by only about 6% because item-level biases shared across samples account for about 87% of squared error.

·3 min
Agents

Compressing old tool outputs cuts coding-agent context with modest accuracy loss

Zou, Wang and Hao introduce a context scheme for software-engineering agents that compresses older tool observations into soft tokens while retaining the agent’s actions and recent observations as text. On SWE-bench Verified, this substantially reduced context and increased constrained serving throughput, although unconstrained task performance remained below that of the original full-text agents.

·2 min