Search Articles — Sudonull

https://sudonull.com/induction-heads-on-tinystories-weak-formation

Experiment shows why TinyStories suppresses Induction Heads in GPT-2. Analysis of SAE, metrics and comparison with Previous Token Heads. Study the impact of the dataset on ICL for your models.

https://sudonull.com/ai-scientific-judge-30b-outperforms-gpt-5-2-in-citability

Learn how Scientific Judge and Thinker models predict arXiv hits from abstracts with 80.6% accuracy. Generalization to domains and self-critique. For AI developers — details of the approach and benchmarks.

https://sudonull.com/gpt-5-4-pro-counterexample-to-gci-for-non-centered-measures

GPT-5.4 Pro disproved the hypothesis of the Gaussian correlation inequality for shifted Gaussian measures. Counterexample in 2D with strips, generalization to n≥2. Solution details on solveall.org. Study for AI in mathematics.

https://sudonull.com/android-app-development-without-experience-ai-case-2500-characters

How a Non-Programmer Created a UTM Generator Using GPT. Technical Nuances, AI Pitfalls, and Lifehacks for Google Play. Learn How to Speed Up Mobile App Development.

https://sudonull.com/ai-releases-composer-2-claude-computer-use

Breakdown of Composer 2 on Kimi, OpenAI purchases, GPT-5.4 Mini, Claude on desktop, Unsloth Studio. Benchmarks, prices, trends for developers. Read reviews and apply in projects.

https://ymaho.com/openai-gpt-rosalind-ai-model-for-faster-drug-research

OpenAI launched GPT-Rosalind to accelerate medical research. Learn how this biology-focused AI works, why access is restricted, and what it means for future treatments.

https://ymaho.com/ai-in-diagnostics-errors-up-to-80-and-gpt-5-hallucinations

Neural networks make mistakes in 80% of cases in differential diagnosis, and GPT-5 produces hallucinations in 65%. Find out why AI is not yet ready to replace a doctor, and where it is truly useful.

https://ymaho.com/thomson-reuters-bought-casetext-for-650-million-strengthening-legal-ai

Thomson Reuters acquires Casetext for $650 million to integrate CoCounsel based on GPT-4. Increasing competition with Bloomberg Law. Deal analysis and forecasts.

https://ymaho.com/sam-altman-s-180m-investment-in-anti-aging-via-ai-and-cells

Analysis of Sam Altman's investments in Retro Biosciences, GPT-4b micro AI model, and cellular reprogramming. Real prospects, risks, and competitors. Read details.

GPT‑2: 1.5B release - OpenAI

https://openai.com/index/gpt-2-1-5b-release/

Nov 5, 2019 · As the final model release of GPT-2’s staged release, we’re releasing the largest version (1.5B parameters) of GPT-2 along with code and model weights to facilitate detection of outputs of …

GPT-2 - Wikipedia

https://en.wikipedia.org/wiki/GPT-2

Generative Pre-trained Transformer 2 (GPT-2) is a large language model (LLM) by OpenAI and the second in their foundational series of GPT models. GPT-2 was pre-trained on a dataset of 8 million …

GitHub - openai/gpt-2: Code for the paper "Language Models are ...

https://github.com/openai/gpt-2

Code and models from the paper "Language Models are Unsupervised Multitask Learners". You can read about GPT-2 and its staged release in our original blog post, 6 month follow-up post, and final …

openai-community/gpt2 · Hugging Face

https://huggingface.co/openai-community/gpt2

Pretrained model on English language using a causal language modeling (CLM) objective. It was introduced in this paper and first released at this page. Disclaimer: The team releasing GPT-2 also …

Introducing ChatGPT Images 2.0 - OpenAI

https://openai.com/index/introducing-chatgpt-images-2-0/

Apr 21, 2026 · ChatGPT Images 2.0 introduces a state-of-the-art image generation model with improved text rendering, multilingual support, and advanced visual reasoning.

ChatGPT

https://chatgpt.com/

ChatGPT is your AI chatbot for everyday use. Chat with the most advanced AI to explore ideas, solve problems, and learn faster.

GPT-2 · Hugging Face

https://huggingface.co/docs/transformers/model_doc/gpt2

Click on the GPT-2 models in the right sidebar for more examples of how to apply GPT-2 to different language tasks. The example below demonstrates how to generate text with Pipeline or the …

Get started | ChatGPT

https://chatgpt.com/auth/login

Log in or sign up to ChatGPT

GPT-2 - OpenAI - LLM Database

https://llmdb.com/models/gpt-2

Initially partially withheld due to concerns about potential misuse, the full model was released in November 2019. Generative Pre-trained Transformer 2 (GPT-2) is a large language model by OpenAI …

GPT-2 — The Complete Guide, (2019–2026) - toolkitbyai.com

https://toolkitbyai.com/gpt-2-complete-guide-2026/

Dec 15, 2025 · GPT-2 is an orthodox transformer-based language model released by OpenAI in 2019. It uses a writer-only architecture and was trained on WebText, a scraped assortment of high-quality …