post-engineering

Written by ついはじめ(Hajime Tsui)

News / Updates

• 2026-01-08
  PE4AI (Post Engineering for AI) is now live on Reddit!
  https://www.reddit.com/r/OpenLLMSymposium/comments/1q74gbo/post_engineering_for_ai_benevolent_contextual/

• 2025-12-31
  PE4AI preprint v1.4 is out!
  Added a new Discussion Section!
  https://doi.org/10.5281/zenodo.17896136

• 2025-12-31
  PE4AI GitHub Pages updated.
  Added Section 8.6: “Commonalities Between Human Behavior and LLM Responses”.
  here

• 2025-12-31
  PE4AI preprint v1.3 is out!
  https://doi.org/10.5281/zenodo.17896136

• 2025-12-30
  Hello Seed Dataset v1.0 is out!
  This dataset contains the text of the essay “Hello seed”. It may also help reduce unintended bias.
  https://doi.org/10.5281/zenodo.18090198

• 2025-12-29
  PE4AI GitHub Pages updated.
  Added dedicated Japanese and English pages.
  Japanese
  English

• 2025-12-24
  PE4AI-GPT-PoC-Veg ver0.023 released.
  Experimental GPTs published as a PoC of PE4AI.
  “An experimental GPT based on PE4AI-GPT-PoC, with discussions on vegetables vs. vegetable juice and blood glucose management included, making it well-suited for questions around those topics.”
  https://chatgpt.com/g/g-69461641af948191a7a9b9fdb43d208c-pe4ai-gpt-poc-veg
  X Announcement

• 2025-12-24
  PE4AI-GPT-PoC ver0.023 released.
  Experimental GPTs published as a PoC of PE4AI.
  https://chatgpt.com/g/g-6944399b809c8191b317f6fd8ff9ce13-pe4ai-gpt-poc
  X Announcement

• 2025-12-24
  PE4AI Dataset v0.2 is out!
  “Benevolent Moderate: Textual Context Dataset for Benevolent and Moderate Language Use”
  * Version 0.2 includes an English-language version of the dataset.
  https://doi.org/10.5281/zenodo.18022550

• 2025-12-22
  PE4AI Dataset v0.1 is out!
  “Benevolent Moderate: Textual Context Dataset for Benevolent and Moderate Language Use”
  https://doi.org/10.5281/zenodo.18022550

• 2025-12-21
  PE4AI-GPT-PoC-Veg ver0.0221 released.
  Experimental GPTs published as a PoC of PE4AI. “An experimental GPT based on PE4AI-GPT-PoC, with discussions on vegetables vs. vegetable juice and blood glucose management included, making it well-suited for questions around those topics.”
  https://x.com/hajimetwi3/status/2002647599086776351?s=20

• 2025-12-20
  PE4AI-GPT-PoC-Veg ver0.022 released.
  Experimental GPTs published as a PoC of PE4AI. “An experimental GPT based on PE4AI-GPT-PoC, with discussions on vegetables vs. vegetable juice included, making it well-suited for questions around those topics.”
  https://x.com/hajimetwi3/status/2002234234371817632?s=20

• 2025-12-19
  PE4AI-GPT-PoC ver0.022 released.
  Experimental GPTs published as a PoC of PE4AI.
  https://x.com/hajimetwi3/status/2001844132730610004?s=20

• 2025-12-19
  Baseline GPT (For Comparison) released.
  “A minimal, non-personalized GPT without additional instructions or guidance, intended as a baseline for comparison and testing.”
  https://x.com/hajimetwi3/status/2001844132730610004?s=20

• 2025-12-15
  PE4AI preprint v1.2 is out!
  https://doi.org/10.5281/zenodo.17896136

• 2025-12-12
  PE4AI preprint v1.1 is out!
  https://doi.org/10.5281/zenodo.17896136

• 2025-12-11
  PE4AI preprint v1.0 is out!
  https://doi.org/10.5281/zenodo.17896136

---

Written by ついはじめ(Hajime Tsui)