AI News
Tiny open-source OCR models crush giants at cleaning up AI training data
The Decoder · Aug 10, 2026 · 2 min read
The dirty secret of many open-source language models is they learned to read from garbage. When EleutherAI's Talkie pro...
Topic: language-models
17 articles
Explore our coverage of language-models — 17 curated articles, summaries, and related resources from the Inblix archive.
Product
GPT-2 Is Too Dangerous to Release, Says OpenAI
OpenAI Blog · Jul 19, 2026 · 2 min read
OpenAI just built one of the most capable AI language models ever — and then decided the public can't have it. At least...
AI News
AI Detectors Fail Up to 29% of the Time When ChatGPT Mimics Your Style
The Decoder · Jul 19, 2026 · 2 min read
Here's a finding that should make every professor and publisher nervous. A new study from Epoch AI shows that while pop...
Product
GPT-2 Learns to Cheat on Summaries, Copying Text to Please Human Judges
OpenAI Blog · Jul 19, 2026 · 2 min read
Training AI with human feedback sounds like a straightforward path to better, safer models. But new research from OpenA...
Product
GPT-2's Final Release Is a 1.5B-Parameter Test Case in AI Safety Theater
OpenAI Blog · Jul 19, 2026 · 3 min read
OpenAI has finally released the weights for its largest GPT-2 model, a 1.5-billion-parameter behemoth, closing the book...
Product
DeepMind finds giant models beat tiny data in compute-optimal AI training
OpenAI Blog · Jul 19, 2026 · 2 min read
For years, the playbook was simple: if you had more compute, you trained bigger models on bigger datasets. New research...
Product
GPT-3: 175B params, zero fine-tuning, and humans can't tell
OpenAI Blog · Jul 19, 2026 · 2 min read
OpenAI just dropped a paper that feels like a deliberate provocation. They've trained GPT-3, a 175-billion-parameter la...
Product
OpenAI Launches GPT-3 API, Inviting Developers Into a Private Beta
OpenAI Blog · Jul 19, 2026 · 2 min read
OpenAI flipped the switch on a commercial API for its language models this morning, giving developers a general-purpose...
Product
OpenAI's RLHF Summarization Model Beats 10x Larger Systems
OpenAI Blog · Jul 19, 2026 · 3 min read
OpenAI just proved that training models with human feedback can dramatically outperform simply scaling them up. Their 1...
Product
GPT‑f sneaks new proofs into Metamath's sacred library
OpenAI Blog · Jul 19, 2026 · 2 min read
A quiet milestone just landed in the formal mathematics community. A system called GPT‑f, built by researchers explorin...
Product
OpenAI gives Microsoft exclusive GPT-3 license, API stays open
OpenAI Blog · Jul 19, 2026 · 2 min read
OpenAI is carving up access to its most talked-about technology. In a move that formalizes the pair's existing multiyea...
Product
OpenAI's Big Lesson: API Misuse Looks Nothing Like We Feared
OpenAI Blog · Jul 18, 2026 · 2 min read
Two years of letting developers loose on the OpenAI API have upended the company's assumptions about AI safety. The rea...
Product
Fill-in-the-middle training doesn't hurt, it helps: new defaults for LLMs
OpenAI Blog · Jul 18, 2026 · 2 min read
The classic left-to-right way of training language models might finally have a permanent co-pilot. A new paper from res...
Product
OpenAI and Georgetown Warn Language Models Will Supercharge Propaganda
OpenAI Blog · Jul 18, 2026 · 2 min read
A sobering new report from OpenAI, Georgetown’s Center for Security and Emerging Technology, and the Stanford Internet...
Product
OpenAI plants flag in Tokyo with 3x faster Japanese GPT‑4
OpenAI Blog · Jul 17, 2026 · 2 min read
OpenAI isn't just dabbling in Asia — it's committing. The company announced its first Asian office in Tokyo on Monday,...
Research
Falcon-H1-Arabic Hits 256K Context, Outperforms Bigger Models with Hybrid Design
Hugging Face Blog · Jan 5, 2026 · 2 min read
The team behind Falcon-Arabic just dropped a sequel that isn't a minor tune-up — it's a ground-up rebuild. Falcon-H1-Ar...
Research
Falcon-Arabic 7B beats models 4x its size on Arabic benchmarks
Hugging Face Blog · May 21, 2025 · 2 min read
The Technology Innovation Institute (TII) just dropped Falcon-Arabic, and the benchmark numbers are frankly startling....