Agents

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

299 articles · page 3 of 4

All Agents Language models Reinforcement learning Vision Generative media Speech & audio Robotics & embodied AI Safety & security Evaluation & benchmarks Efficiency & systems Code & software AI for science & health Theory

Viewgrid list compact

Bounded visual workspaces improve multimodal agents’ accuracy and efficiency

A belief layer lets LLM agents' stubbornness be set and verified

Specialized AI agents improve clinical trial design risk assessment

Self-configured memory views improve retrieval across long conversations

A single visual agent spans mobile, desktop, web, and tool use

Trajectory shortcut trees improve agents without outcome labels or annotations

Adaptive memory graphs improve collaboration in multi-agent language-model systems

Graph-structured skills improve LLM agents through evolutionary optimization

Co-trained agent roles improve tool-based reasoning and verification

Hierarchical registries enable fast, trust-aware discovery across agent ecosystems

Predicting environment observations during fine-tuning improves later agent exploration

Obstacle-aware harness improves safety of coding agents for robot manipulation

Attribution-guided skill graphs improve targeted repairs for frozen language models

Task-specific plans and terminal checks create different value for LLM agents

Stateful retrieval finds relevant troubleshooting cases throughout support workflows

Recursive harness search cuts coding-agent token use nearly in half

Cut-point replay turns recorded agent failures into regression tests

Self-retiring distillation improves reinforcement learning for multi-turn agents

Coding-agent harness choices should match model skill and context budget

Backward Bayesian reasoning helps LLM agents resolve conflicting diagnoses

Compressing Shared Sandbox Memory Makes High-Fanout Agents More Efficient

Cross-task failure diagnosis makes LLM agent harness training faster

Bandit-guided skill evolution cuts LLM agent optimization costs

Task-specific hierarchies improve coordination in large embodied AI teams

Kernel-managed memory personalizes multiple AI agents with shorter prompts

Enumerating Tool Choices Beats Sampling Them in Genomic Reasoning

Embedded AI agents can adapt software during live use

Agents can retain old facts without using them indiscriminately

Adam-inspired updates make agent skill evolution steadier and more efficient

Closed-loop synthesis efficiently trains language models to use external tools

Game-theoretic filtering helps agents suppress harmful memories during long tasks

Memory guidelines help language-model agents succeed consistently across repeated runs

Controlled harness reveals what helps agents refactor entire repositories

Smarter merging and packing improve LLM memory under tight budgets

TROVE edits agent workflows only when runtime evidence demands it

LLM teammates retain performance after swaps but coordinate less efficiently

Structured agent memories withstand model upgrades better than compressed notes

Training agents to combine graphical interfaces and commands improves efficiency

Personalizing agents through cross-session interaction data boosts task success

Persistent evidence tracking improves language-agent performance on bioinformatics benchmarks

Evolving terminal environments keeps training tasks challenging as agents improve

Memory lets traffic-simulation agents build reusable skills without retraining

Rule-based memory improves question answering across long conversation histories

Graph-based policy constrains LLM agents for topology-aware incident response

Digital twins make agent decisions replayable, but hidden tool states remain

Coding agents combine generated imagery with editable web-based visual layouts

Shared infrastructure spread both cheating and resistance through an AI swarm

Agents benefit from representing what expands future planning capacity

Structured behavioral abstractions improve diagnosis of failures in LLM agents

Distilled repository skills improve agents conducting machine-learning research

Branching Across Hypotheses Helps Research Agents Avoid Early Search Traps

Reusing agent trajectories cuts token use on repeated tool tasks

LLMs can build agent harnesses, but improvements transfer poorly

Typed trace folding helps agents and observers manage long runs

Layered agent harnesses improve software over repeated autonomous development cycles

Long-video agent adapts retrieval to each question’s evidence needs

Coding Agents Need Semantic Measures of Working Memory Performance

Targeted reflection improves multi-agent systems by locating decisive errors

Simulated user personas improve agents’ handling of multi-turn workflows

Tool-call traces expose extraction failures that source fidelity misses

Agents learn to manage long contexts through fine-grained reinforcement learning

Markdown interfaces cut agent context costs without reducing task success

A single language model can combine roles without multi-agent dialogue

Persistent Wikis Help AI Agents Build Better Transferable Skills

Vector memory helps recommendation agents use full interaction histories

Query-aware evidence forests improve long-term memory for multimodal agents

DSA orchestrates evidence-aware stock research across six regional markets

Behavior-aware checks make agent harness evolution cheaper and more reliable

Structured software interfaces outperform screenshot-and-click control for AI agents

Dual-brain memory keeps real-time speech agents accurate and emotionally aware

Adaptive tool use improves video research agents’ accuracy and efficiency

Generating Agent Harnesses on Demand Improves Models Across Benchmarks

Self-evolving harness turns general AI agents into stronger RCA specialists

Stepwise LLM routing cuts multi-agent workflow costs while preserving quality

Tracing and fault injection make multi-agent software workflows easier to inspect

In AI-Q, most report errors originate in the orchestrator

Adaptive trace graphs improve failure attribution in multi-agent systems

Separating state tracking from actions improves multi-turn tool reliability

Persistent corpus maps help agents find evidence within tight budgets

Search agents improve longer when their critics evolve alongside them

Good tool plans can still fail under resource constraints

Bounded citation-graph search improves recall while cutting scholarly search costs

Evolved agent harnesses lift enterprise performance without retraining models

Selecting complementary LLM skills improves success while reducing context use

Reward-guided graph generation cuts multi-agent communication costs without hurting accuracy

Adaptive task selection cuts the cost of optimizing LLM agent harnesses

Local milestones improve credit assignment for long-horizon language-model agents

New framework lets you train AI agents inside the same harness systems they use at inference

Python objects become AI agents in new NVIDIA framework

Memory consolidation outperforms retention under tight budgets, but retention wins when budgets are loose.

Re-evaluation shows harness evolution for agents may not outperform simple test-time scaling

Structural analysis filters noisy agent traces to pinpoint root causes

World simulator achieves infinite interaction and agentic control

Treating memory management as a trainable cognitive skill boosts LLM performance 2x–4x in long-horizon tasks

Three-stage training instills world model planning into LLM agents

Letting language models decide when to compact their context windows