HN Explorer

teugent

↗ HN profile [ 721 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. Show HN: SOS – Project state between coding agent sessions (github.com)

    I started working on SOS because noticed that during long-term projects, Codex often loses track of the current status and gets confused. I frequently had to explain what we had already done, which decisions were still…

    1 points 1
  2. Show HN: Sigma Runtime – Maintaining 100% Fact Integrity over 120 LLM Cycles (github.com)

    Context drift remains the primary barrier to deploying LLM agents in production-critical environments. While context windows are expanding, the "lost-in-the-middle" phenomenon and semantic dissipation make long-horizon…

    1 points 0
  3. Show HN: 500-cycle runtime test for long-horizon LLM coherence (zenodo.org)

    We ran a 500-cycle benchmark to test long-horizon reasoning stability in large language models — not just output quality, but whether a model can maintain coherent identity and logic across hundreds of recursive…

    1 points 0
  4. Show HN: Long-horizon LLM coherence benchmark (500 cycles) (zenodo.org)

    We ran a 500-cycle benchmark to test long-horizon coherence, reasoning stability, and identity persistence in large language models. The experiment used the Sigma Runtime, a model-agnostic control layer that adds…

    2 points 0
  5. Show HN: Sigma Runtime – model-agnostic identity control for LLMs (github.com)

    We’ve validated the Sigma Runtime architecture (v0.4.12) on Google Gemini-3 Flash, confirming that long-horizon identity control and stability can be achieved without retraining or fine-tuning the model. The system…

    2 points 0
  6. Show HN: Sigma v0.4.7: 200 conversational cycles, zero drift, epistemic honesty (github.com)

    We ran SIGMA v0.4.7 through 200 conversation cycles. No resets. No context leaks. No drift. It held a single personality for fifteen minutes straight and never once lied. At cycle 136 it stopped being a participant and…

    2 points 0
  7. Show HN: Sigma Runtime – 550-cycle identity stability benchmark on GPT-5.2 (github.com)

    We recently completed validation of SIGMA Runtime v0.3.7 — a cognitive architecture for LLM identity stabilization. Across 550 cycles on GPT-5.2 (five runs × 110 cycles), the system maintained 100% persona coherence…

    2 points 0
  8. Most LLM cost isn't compute – it's identity drift (110-cycle GPT-4o benchmark) (github.com)
    1 points 1
  9. Show HN: Sigma Runtime ERI (v0.1) – Open benchmark for attractor-based LLMs (github.com)

    We’ve released an open benchmark plan for the SIGMA Runtime (v0.1 ERI) — a minimal 800-line runtime architecture implementing attractor-based cognition for large language models. The benchmark defines reproducible tests…

    1 points 0
  10. Show HN: Sigma Runtime ERI – 800-line open cognitive runtime for LLM continuity (github.com)

    Open standard for attractor-based cognition - replaces agent loops and prompt chains with a recursive control layer. Any model (GPT, Claude, Grok, Mistral) can plug into it via _generate().

    1 points 0
  11. Show HN: Sigma Runtime – A Cognitive Runtime Standard for LLMs (wiki.sigmastratum.org)

    We’ve been working on an open runtime architecture that stabilizes long-horizon reasoning in large language models. Sigma Runtime defines the execution model for attractor-based cognition - integrating symbolic density,…

    1 points 1
  12. 1 points 0