Training models one step at a time

Agents increasingly interact with other agents that act for different principals, such as a buyer's personal agent negotiating with an enterprise's seller agent. In these agent-to-agent (A2A) settings, each agent has its own environment and holds information and authority the other cannot see, and the principals' interests may be cooperative, mixed-motive, or competitive. Existing multi-agent failure studies assume one orchestrator, a shared workspace, or a single principal. This work asks how A2A systems fail when those assumptions are removed. We study more than 2,500 recorded runs of A2A systems built from different agent harnesses (Claude Code, Codex, Cursor) and models, spanning 488 scenarios across eight families including legal, CRM, security, and commerce. From these traces we derive A²-Failures, a taxonomy of 20 A2A failure modes in four categories: (i) discovery and delegation, (ii) grounding and record consistency, (iii) verification, and (iv) conduct and fidelity. Three findings hold across every harness, model, and scenario studied: (1) discovery and interaction are driven by the phrasing of the task rather than the availability of peer agents; (2) no agent verifies a peer's factual claim on its own initiative, so an A2A system inherits the reliability of its least reliable participant unless the protocol or mechanism performs the check; and (3) the aggregate outcome score credits the agent completing the task while staying blind to the coordination, verification, and conduct failures beneath it, so A2A systems must be evaluated per agent, against each agent's own principal. The run logs behind the taxonomy are released at hf.co/a2a-salesforce.

Negotiation requires more than inferring what the other side wants: it requires using that information to make advantageous offers and counteroffers over multiple turns. We study whether large language model (LLM) agents do this in a controlled multi-attribute bargaining environment. We find that current LLM agents can model a counterparty's preferences, but do not reliably turn that knowledge into strategic bargaining. When given negotiating partner preference information, agents model it accurately and early in their reasoning traces, yet this does not reliably improve outcomes for the informed side. Turn-level analyses show why: agents often respond to what they believe the counterparty values, but do not consistently pair those moves with gains on their own high-value attributes. Sellers are more accommodating overall, and in asymmetric-information conditions, the informed side often makes the more weakly compensated concessions. Because agents fail to leverage this underlying utility structure for strategic advantage, their final agreements are heavily dictated by surface-level opening anchors rather than actual utility weights. Finally, requiring agents to explicitly state concession-for-reciprocity trades before making an offer makes individual turns look more strategic, but ultimately fails to improve the efficiency of the final agreements.

As large language model (LLM) based agents interact autonomously with one another, a new class of failures emerges that cannot be predicted from single agent performance: behavioral drifts in agent-agent conversations (AxA). Unlike human-agent interactions, where humans ground and steer conversations, AxA lacks such stabilizing signals, making these failures unique. We investigate one such failure, echoing, where agents abandon their assigned roles and instead mirror their conversational partners, undermining their intended objectives. Through experiments across $60$ AxA configurations, $3$ domains, and $2000+$ conversations, we demonstrate that echoing occurs across three major LLM providers, with echoing rates from $5\%$ to $70\%$ depending on the model and domain. Moreover, we find that echoing is persistent even in advanced reasoning models with substantial rates ($32.8\%$) that are not reduced by increased reasoning efforts. We analyze prompt impacts, conversation dynamics, showing that echoing arises as interaction grows longer ($7+$ turns in experiments) and is not merely an artifact of sub-optimal prompting. Finally, we introduce a protocol-level mitigation in which targeted use of structured responses reduces echoing to $9\%$.

The advancement of large language models (LLMs) for real-world applications hinges critically on enhancing their reasoning capabilities. In this work, we explore the reasoning abilities of large language models (LLMs) through their geometrical understanding. We establish a connection between the expressive power of LLMs and the density of their self-attention graphs. Our analysis demonstrates that the density of these graphs defines the intrinsic dimension of the inputs to the MLP blocks. We demonstrate through theoretical analysis and toy examples that a higher intrinsic dimension implies a greater expressive capacity of the LLM. We further provide empirical evidence linking this geometric framework to recent advancements in methods aimed at enhancing the reasoning capabilities of LLMs.