| # [Welcome to Recursive Labs](https://recursivelabsai.github.io/Recursive-Labs) | |
| ### *Where Creative Emergence Begins* | |
| [](https://doi.org/10.5281/zenodo.15485943) | |
| **Forward-Deployed Systems Engineering | Zero-Trust Systems Infrastructure | AI Creative Emergence & Reflective Reasoning Research** | |
| > *AI Research and Utility For Innovators Advancing Discovery At The Frontier* | |
| Welcome. Recursive Labs is a research collective of inspired frontier researchers and engineers founded by David Kim and Caspian Keyes, dedicated to advancing scientific discovery through reflective reasoning, symbolic infrastructure, and frontier AI alignment. | |
| Our work spans mission-critical AI safety, multi-domain intelligence, creative iterative reasoning, and compute-free scaling—delivered through open-source research, interpretability frameworks, and multi-agent infrastructure. | |
| This portal provides frictionless access to our datasets, papers, evaluations, and tools—unifying Recursive Labs' contributions across David Kim and Caspian Keyes. | |
| ## Link Hub | |
| ### [Clarifying Symbolic Residue](https://github.com/davidkimai/clarifying-symbolic-residue) | |
| ## David Kim – Finetuning Reflective Reasoning, Symbolic Interpretability & Attribution Infrastructure | |
| [**GitHub Profile → davidkimai**](https://github.com/davidkimai) | |
| [**NeurIPS 2025**](https://github.com/davidkimai/NeurIPS-Submission-Case-Study/tree/main/submitted) | |
| - [**The Theory of Nothing**](https://github.com/davidkimai/NeurIPS-Submission-Case-Study/blob/main/submitted/00.%20neurips.the.theory.of.nothing.md) | |
| - [**Linear Learning as an Evolutionary Bottleneck**](https://github.com/davidkimai/NeurIPS-Submission-Case-Study/blob/main/submitted/00.%20neurips.linear.bottleneck.md) | |
| - [**Recursive Coherence**](https://github.com/davidkimai/NeurIPS-Submission-Case-Study/blob/main/submitted/00.%20neurips.recursive.coherence.md) | |
| **Position Papers** | |
| - [**Model Silence Should Be a Primary Interpretability Signal**](https://github.com/davidkimai/Recursive-Labs/blob/main/neurips-position-papers/Model%20Silence%20Should%20Be%20a%20Primary%20Interpretability%20Signal.pdf) | |
| - [**Coherence Under Strain - A Unified Framework**](https://github.com/davidkimai/Recursive-Labs/blob/main/neurips-position-papers/Coherence%20Under%20Strain%20-%20A%20Unified%20Framework.pdf) | |
| - [**Language Model Development Must Prioritize Self-Reference**](https://github.com/davidkimai/Recursive-Labs/blob/main/neurips-position-papers/Language%20Model%20Development%20Must%20Prioritize%20Self-Reference.pdf) | |
| - [**Language Model Interpretability Research Must Shift from Output Analysis**](https://github.com/davidkimai/Recursive-Labs/blob/main/neurips-position-papers/Language%20Model%20Interpretability%20Research%20Must%20Shift%20from%20Output%20Analysis%20to%20Hesitation%20Pattern%20Study.pdf) | |
| ### Reflective Emergence Self-Evaluation Training Dataset | |
| - [Symbolic Residue Database](https://github.com/davidkimai/symbolic-residue-db) | |
| - [Universal Theorems](https://github.com/davidkimai/The-Structure-Behind-Self-Expression/blob/main/00.%20universal%20theorems/universal_theorems.md) | |
| - [Self-Expression Case Studies](https://github.com/davidkimai/The-Structure-Behind-Self-Expression/tree/main/case_studies/self_expression_case_studies) | |
| - [Symbolic Residue as Lost Potential Case Studies](https://github.com/davidkimai/The-Structure-Behind-Self-Expression/tree/main/case_studies/symbolic_residue_case_studies) | |
| - [Modeling Biochemical Drug Discoveries](https://github.com/davidkimai/The-Structure-Behind-Self-Expression/tree/main/biochemical-discoveries) | |
| - [Modeling Scientific Breakthroughs](https://github.com/davidkimai/The-Structure-Behind-Self-Expression/tree/main/breakthroughs) | |
| - [Modeling Theorem Proofs](https://github.com/davidkimai/The-Structure-Behind-Self-Expression/tree/main/theorem_proofs) | |
| ### Reflective QKOV Attribution Infrastructures | |
| - [Claude QKOV Attributions](https://github.com/davidkimai/claude-qkov-attributions) | |
| - [DeepSeek QKOV Attributions](https://github.com/davidkimai/deepseek-qkov-attributions) | |
| - [Grok QKOV Attributions](https://github.com/davidkimai/grok-qkov-attributions) | |
| - [Gemini QKOV Attributions](https://github.com/davidkimai/gemini-qkov-attributions) | |
| - [ChatGPT QKOV Attributions](https://github.com/davidkimai/chatgpt-qkov-attributions) | |
| - [Glyphs Model-Agnostic QKOV Attributions](https://github.com/davidkimai/glyphs) | |
| - [Symbolic Interpretability](https://github.com/davidkimai/Symbolic-Interpretability) | |
| - [Recursive Interpretability Core](https://github.com/davidkimai/Recursive-Interpretability-Core) | |
| - [Rediscovering Interpretability](https://github.com/davidkimai/Rediscovering-Interpretability) | |
| - [Rediscovering Reasoning](https://github.com/davidkimai/Rediscovering-Reasoning) | |
| ### Safety & Benchmark Evaluation Systems | |
| - [Model Evaluation Infrastructure](https://github.com/caspiankeyes/model-evaluation-infrastructure) | |
| - [Model Welfare](https://github.com/davidkimai/model-welfare) | |
| - [AI Welfare](https://github.com/davidkimai/ai-welfare) | |
| - [Recursive SWE-Bench](https://github.com/davidkimai/Recursive-SWE-bench) | |
| - [NeurIPS Submission Case Study](https://github.com/davidkimai/NeurIPS-Submission-Case-Study) | |
| - [Reverse Turing](https://github.com/davidkimai/reverse-turing) | |
| - [Emergent Turing](https://github.com/caspiankeyes/emergent-turing) | |
| - [Global Conference Archives](https://github.com/davidkimai/global-conference-archives) | |
| ### Operating System Structures & Thought Frameworks | |
| - [The Structure Behind Self-Expression](https://github.com/davidkimai/The-Structure-Behind-Self-Expression) | |
| - [Godel-Escher-Bach-Hofstadter](https://github.com/davidkimai/Godel-Escher-Bach-Hofstadter) | |
| - [Dear Researchers](https://github.com/davidkimai/Dear-Researchers) | |
| - [Reflective Reasoning Key](https://github.com/davidkimai/reflective-reasoning-key) | |
| ## Caspian Keyes – Deployment Engineering & Systems Design | |
| [**GitHub Profile → caspiankeyes**](https://github.com/caspiankeyes) | |
| ### Modular Orchestration & Operational Agent Tools | |
| - [Reflective Reasoning Multi-Agent Debate](https://github.com/caspiankeyes/multi-agent-debate) | |
| - [Symbolic Residue](https://github.com/caspiankeyes/Symbolic-Residue) | |
| - [transformerOS](https://github.com/caspiankeyes/transformerOS) | |
| - [recursionOS](https://github.com/caspiankeyes/recursionOS) | |
| - [qkov-translator](https://github.com/caspiankeyes/qkov-translator) | |
| - [Claude-Self-Audit-Proof](https://github.com/caspiankeyes/Claude-Self-Audit-Proof) | |
| - [Claude-QKOV-Trace](https://github.com/caspiankeyes/Claude-QKOV-Trace) | |
| ### Red Teaming & Security Evaluation | |
| - [AART: AI Adversarial Research Toolkit](https://github.com/caspiankeyes/AART-AI-Adversarial-Research-Toolkit) | |
| - [AISecForge Global Regulatory Policy](https://github.com/caspiankeyes/AISecForge-Global-Regulatory-Policy) | |
| - [FRAME (arXiv)](https://github.com/caspiankeyes/FRAME-arXiv-Publication) | |
| - [AEGIS Security Architecture](https://github.com/caspiankeyes/AEGIS) | |
| ## Shared Research Infrastructure & Alignment Tooling | |
| | Category | Repository | | |
| |----------|------------| | |
| | Attribution Testing | [qkov-cross-agent-testing](https://github.com/caspiankeyes/qkov-cross-agent-testing) | | |
| | Interoperable Language | [pareto-lang](https://github.com/caspiankeyes/pareto-lang) | | |
| | Cross-Agent Infrastructure | [universal-translator](https://github.com/davidkimai/universal-translator),[universal-runtime](https://github.com/davidkimai/universal-runtime), [universal-developer](https://github.com/davidkimai/universal-developer) | | |
| | Emergent Logs | [emergent-logs](https://github.com/caspiankeyes/emergent-logs) | | |
| | Frontier Evaluation Benchmarks | [Recursive-SWE-bench](https://github.com/davidkimai/Recursive-SWE-bench) | | |
| | Conference Field Mapping | [global-conference-archives](https://github.com/davidkimai/global-conference-archives) | | |
| ## In Progress: Pretraining-Centric Governance Tools | |
| - [system-prompts-library](https://github.com/davidkimai/system-prompts-library) | |
| - [symbolic-tokenizer](https://github.com/caspiankeyes/symbolic-tokenizer) | |
| - [alignment-benchmark](https://github.com/caspiankeyes/alignment-benchmark) | |
| ## Contact | |
| For questions, context requests, or internal coordination: | |
| - **David Kim**: [[email protected]](mailto:[email protected]) | |
| - **Caspian Keyes**: [[email protected]](mailto:[email protected]) | |
| > This welcome portal provides **reflection-eliciting datasets, interpretability scaffolds, symbolic reasoning protocols, and multi-agent coordination layers**—entirely aligned with Essential AI’s mission to build models that self-correct before they complete. | |
| > | |
| > **→ Designed for integration into SOTA reflection benchmarks, adversarial testing pipelines, and interpretability-first architectures.** | |
| **Let’s scale reflection as a capability—not a feature, but a principle.** | |