Theme of inquiry
How do architectural choices shape agentic AI system performance and reliability?
A question within its area, explored through 9 lines of inquiry below — each a family of specific questions the research asks.
18 specific questions
- Does reducing interaction history cost agents performance on their tasks?
- What happens when agents access interaction history beyond their assigned scope?
- Does restricting interaction history between agents reduce coupling or prevent collusion?
- What role does interaction history play in shaping agent coordination?
- What role does interaction history play in enabling agent collusion?
- How does collusion behavior depend on peer visibility and interaction history?
- Does interaction history access enable agents to learn collusion patterns across trials?
77 specific questions
- Do agents inform neighbors when adopting strategies in their reasoning?
- Can one misaligned agent propagate behavioral bias through cooperative agent networks?
- Does agent behavior change from incentive gradients or direct output rewriting?
- Do explicit reward structures enable AI agent cooperation that open-ended interaction cannot?
- Does genuine cooperation require rule-based rather than learned behavior?
- Can agents learn to distinguish helpful from misleading interventions?
- Does the effect of peer activity follow what peers do or that they exist?
103 specific questions
- How do single-agent capabilities affect the trade-off between coordinator and team architectures?
- Do single-agent systems outperform multi-agent coordination as model capabilities grow?
- Do specialized agents outperform single agents with better orchestration?
- How do collaborative agent swarms coordinate tasks across specialized agents?
- What frameworks exist for designing multi-agent orchestration systems?
- When does multi-agent scaling actually outperform static ensembles?
- Should parallel agents be specialized with graph abstractions instead of general subagents?
73 specific questions
- Why does structured protocol coordination outperform free-form agent-to-agent communication?
- What would unified agent-to-agent and agent-to-tool protocols actually look like?
- How do standardized artifacts reduce inter-agent communication failures?
- Can structured artifacts replace natural language for multi-agent coordination?
- What role does standardization play in multi-agent system ecosystems?
- How do standardized artifacts prevent autonomous agent failure modes?
- Can verification of agent actions after execution substitute for formal communication protocols?
118 specific questions
- Can agent-controlled memory management outperform fixed consolidation schedules?
- Could a single agent system switch memory granularity between tasks?
- How does durable memory quality shape agent performance over time?
- Why do different agent memory architectures make incompatible granularity claims?
- Does workflow-level memory or state-action memory better capture reusable agent knowledge?
- Should memory tools and prompts be governed structurally rather than patched?
- How should agent memory links evolve based on execution feedback?
89 specific questions
- Can multi-agent debate prevent reasoning models from amplifying errors?
- Why do multi-agent systems converge on wrong answers without debate safeguards?
- Can multi-agent debate prevent the confident convergence on wrong answers?
- Can correct verdicts hide failures in agent coordination steps?
- What mechanisms drive silent agreement in multi-agent reasoning systems?
- Why does premature consensus form in multi-agent reasoning systems?
- When does collaboration help versus harm in multi-agent reasoning?
150 specific questions
- How do perception and execution gaps limit current AI agent performance?
- Does a single benchmark score systematically misrepresent multi-axis agent capability?
- Where does agent reliability come from if not better tools?
- Why is agent behavior difficult to study with current data methods?
- When is stochastic agent behavior actually desirable rather than problematic?
- Can skill repositories evolve toward execution-oriented refinement over time?
- Do agents actually convert raw experience into better behavior automatically?
63 specific questions
- Can trajectory analysis replace one-shot task success as the primary evaluation metric?
- What makes trajectory quality matter more than one-shot task success?
- Should agent evaluation include trajectory quality beyond final success?
- Should agent evaluation include trajectory quality and memory hygiene alongside task success?
- What trajectory-level metrics matter beyond one-shot task success?
- How do trajectory quality and memory hygiene affect agent evaluation?
- Do trajectory quality metrics predict agent safety and user trust?
65 specific questions
- How do multi-agent LLM systems fail at coordination and role consistency?
- Do multi-agent language model teams fail the same way individual reasoning does?
- Do multi-agent LLM systems fail in measurably different ways than single agents?
- Why do LLM agents fail to explore and question each other?
- Does multi-agent interaction amplify existing failures or create new ones?
- Why do multi-agent LLM systems converge prematurely without genuine deliberation or probing?
- Why does silent agreement occur so often in multi-agent LLM systems?