The paper introduces Contrastive Attribution Topology Search (CATS), a method for selecting compact, high‑performing teams of agents within a larger multi‑agent system. CATS uses an LLM attributer to assign credit to agents based on execution traces, then iteratively prunes weak edges and agents to form a small, efficient team. Experiments show CATS improves accuracy on BigCodeBench by 5.3 percentage points over heuristic search while reducing deployment tokens by 54%, matches strong baselines on MBPP with fewer tokens, and successfully removes adversarial agents in a HotpotQA stress test.
arXiv:2605. 25815v4 Announce Type: replace Abstract: Agent-to-Agent (A2A) networks enable autonomous AI agents to collaborate by sharing reusable problem-solving instructions.
arXiv:2609. 37968v1 Announce Type: new Abstract: Advances in the coding capabilities of LLM agents allow them to inspect and modify their own instructions, tools, and execution procedures.
The paper introduces Codebook Agent, a method for adapting the communication topology of large language model (LLM) multi‑agent systems to individual queries. Instead of searching the full adjacency space, it compresses successful topologies into a 16‑entry codebook via a vector‑quantized autoencoder and uses a reward‑weighted MLP to select a code for each query, followed by a single‑pass MLP proxy to rerank candidates. This approach eliminates iterative search and message passing at test time, achieving higher accuracy across six benchmarks, faster topology generation (2.4 ms), and reduced token usage (21.9–33.2 %).
BekchiAI introduces a benchmark and platform for evaluating large language model agents. The benchmark comprises 13 tool‑using ReAct agents across seven task categories, totaling 2,057 deterministic test tasks with verifier‑checkable gold answers. The platform offers web‑based observability, token and latency telemetry, and remote run termination for live agents.
arXiv:2606. 15899v1 Announce Type: cross Abstract: Open-source LLM agent ecosystems are growing rapidly, yet the security of community-contributed skills - modular tool definitions that extend agent capabilities - remains largely unvetted.
arXiv:2606. 31174v1 Announce Type: new Abstract: Production large language-model (LLM) agents are increasingly deployed not as lone problem-solvers but as managers: a main model creates specialized subagents, delegates work, and orchestrates their parallel, asynchronous returns through dynamic workflows.
The paper introduces BenchAgent, a framework for evaluating multi‑agent large language model (LLM) workflows against a single‑agent baseline, BenchAgent Core. Using GPT‑4.1 across ten reasoning, coding, and tool‑use benchmarks, the study compares six multi‑agent configurations, finding that EvoAgent achieves the highest mean pooled accuracy (79.70%) while other agents lag by 1.71–8.76 percentage points. The analysis also highlights trade‑offs in token consumption and accuracy, offering guidance for selecting and optimizing LLM multi‑agent workflows.
arXiv:2607. 24063v1 Announce Type: new Abstract: On standard factuality tasks, frontier models now cluster near the top of the scale.
The paper introduces a two‑phase pipeline that mines reproducible design principles for large language model multi‑agent systems by reusing all intermediate search trajectories from an MCTS optimizer. Using six composable operators and 26 Enigmata tasks, the authors generate a corpus of 28,568 executable workflows and apply an AutoResearch loop to discover and validate orchestration patterns. Key findings include that Star‑first workflows underperform, SharedMemory outperforms debate, and prompt specificity and placeholder‑based routing provide comparable gains to communication‑structure selection.
arXiv:2608. 16801v1 Announce Type: new Abstract: We study how teams of AI coding agents coordinate while solving programming tasks.
The study compares single-agent and multi-agent large language model (LLM) setups for reviewing credit assessment memos against specified requirements. In a controlled experiment with 300 synthetic memos, a single agent examined six requirements in one call, while a team of three agents each checked two requirements. Results show the team identified 0.08 to 0.10 more defects on certain models but also produced more false alarms, used 2.2 to 2.4 times more tokens, and had mixed effects on balanced accuracy depending on instruction and model.
arXiv:2606. 19704v1 Announce Type: new Abstract: Agent benchmarks are growing fast, but no single benchmark touches more than four or five of the dimensions that deployment exposes.