Research on 00n.ai | Architecture & Research

Agents, Objectives, and Proxies

Sat, 20 Jun 2026 00:00:00 +0000

A working section on what an agent is, why agentic systems need objectives, and why proxies are needed for measurement.

Definitions

Sat, 20 Jun 2026 00:00:00 +0000

Canonical working definitions for the research site.

Evidence-Weighted Routing and Error Measurement for Code Localization and Agentic Repair: A Three-Phase Study

Thu, 04 Jun 2026 00:00:00 +0000

Abstract

This paper studies whether a hierarchical evidence tree, coupled with local support/residual measurements, can improve code localization and downstream coding-agent repair. The work is organized as a single experimental program with three phases: (i) construction of a grounded repository-state tree for traceability and localization, (ii) reformulation of traceability drift as evidence-weighted innovation over observable state rather than hyperbolic transport geometry, and (iii) a frozen-edit Codex augmentation replay to test whether routing and evidence packets reduce wasted code and wrong-region edits.

Designing Multi-Agent Systems

Fri, 19 Jun 2026 00:00:00 +0000

Notes on roles, coordination, memory, routing, and evaluation for reliable agent teams.

Coding Knowledge Graph Agent Benchmark

Fri, 26 Jun 2026 00:00:00 +0000

Does a code knowledge graph with multi-agent navigation help small models write correct code? 4 models × 4 conditions × 9 tasks × 5 runs.

Knowledge Graph Agent Benchmark

Thu, 25 Jun 2026 00:00:00 +0000

Can a reasoning graph close the gap between small (7-8B) and frontier LLMs on statutory reasoning? Multi-agent, saturation, and synthesis scaffolding experiments.

Related Research: Multi-Agent Framework Benchmark Validation

Wed, 24 Jun 2026 00:00:00 +0000

Mapping our multi-agent benchmark findings to verified papers from arXiv, ACL, NeurIPS, ICML, and EMNLP. 17 papers evaluated for support, contradiction, and nuance.

Multi-Agent Framework Benchmark

Tue, 23 Jun 2026 00:00:00 +0000

Comparison data for vanilla, search, reflection, and multi-agent answer pipelines on freshness questions.

iTrust Requirement-to-Code Traceability Benchmark

Sat, 13 Jun 2026 00:00:00 +0000

Canonical registry entry for the iTrust-backed requirement-to-code traceability benchmark package.