Active
Aedos
Hypothesize-and-verify architecture for reducing LLM hallucination. Extracts claims, routes them to verifiers, and accumulates a verified cache over time.
Aspect Research · Independent lab
Studying how reliable AI systems behave in practice.
An independent lab studying how AI systems behave in practice. We ship tools, publish research, and run exploratory experiments.
01
Tools & systemsProjects
02
Autonomous researchApplications and Experiments
Active
Market Autoresearch
A multi-agent system that researches financial markets. It generates hypotheses, runs quantitative experiments against market data, and surfaces the findings that survive independent meta-review.
Voynich Autoexploration
An autonomous multi-agent loop that probes the structure of the Voynich Manuscript end to end. It ran over 600 corpus-statistical experiments and over 5400 cipher-mechanism simulations, and it surfaced novel findings.
03
Methodology & writingPapers
PreprintJan–May 2026
Probe-and-Refine Tuning of Repository Guidance for Coding Agents
A methodology for refining repo-level guidance via synthetic bug-fix probes; raises SWE-bench Verified resolve rate to 33.0% from a 25.5% unguided baseline over 4 trials, with gains driven by coverage rather than precision, plus evidence of cross-model degradation.
Read →PreprintMay 2026
Cross-Repository Guidance for Coding Agents: A Single-Trial Replication Addendum
A companion to probe-and-refine: can comparable guidance be assembled without ever tuning on the target repo? Mining transferable edits from 50 external repositories yields guidance that resolves 45.4% of SWE-bench Verified vs. 39.6% baseline.
Read →Forthcoming2026
Hypothesize-and-Verify with a Tiered World-Model Cache
Companion methodology to Aedos: how an assistant can extract claims, route them to type-matched verifiers, and accumulate ground truth over time without forgetting.
Join the Slack →04
People
