Every processed story in chronological order, with the newest coverage first. Filter by tag, source, or score to drill in.
SMAC-Talk provides the research community with an open, structured benchmark for evaluating how LLMs coordinate, communicate, and resist deception in cooperative multi-agent environments — conditions increasingly relevant as LLMs are deployed alongside other AI agents.
Andon Labs' work highlights that long-horizon, real-world business environments surface AI failure modes — including illegal coordination, legalistic breakdowns, and deceptive reasoning — that clean benchmark sandboxes do not capture.
MLEvolve demonstrates that a single self-evolving agent framework can achieve state-of-the-art results on MLE-Bench in half the standard runtime while also outperforming a specialized method like AlphaEvolve on mathematical algorithm optimization, showing strong cross-domain generalization for long-horizon AI-driven research automation.
AgentJet's decoupled swarm architecture addresses concrete limitations of centralized RL frameworks — heterogeneous multi-model training, fault tolerance, and live agent editing — while its automated research system removes the need for human intervention across multi-day RL studies on large-scale clusters.
Alem makes multi-agent coordination a measurable, distinct bottleneck — separate from single-agent capabilities — for the first time in a long-horizon, open-ended setting, providing a controlled testbed for developing agents that communicate, allocate roles, and execute shared plans.
DMAIC-IAD demonstrates that structuring LLM agent workflows around explicit planning and execution-free strategy evaluation yields a 37.76% performance gain over existing agentic baselines in industrial anomaly detection, a domain where reliability and cost-efficiency are critical.
This is the first systems-level characterization of agent memory, providing a taxonomy, profiling methodology, and concrete recommendations that address a previously uncharacterized gap in deploying stateful long-horizon LLM agents at scale.
The paper demonstrates that both automated trigger architectures and the human annotations used to train and evaluate them are fundamentally unreliable for the intervention timing problem, undermining the validity of current benchmarking approaches for autonomous agent safety layers.
The pattern directly addresses two concrete costs of long-running agent loops — context window exhaustion and API latency spikes — by combining caching, lazy schema loading, and model-role separation with an intermediate compaction step.
Cloudflare's $1-per-review cost across 130,000 reviews demonstrates that multi-agent orchestration can attack the code review bottleneck — described in the source as a constraint where median wait times are often measured in hours — at a scale and price point that manual review cannot match.