Pangram Clarified: Not a Measure of Writing Quality
TuhinChakr · x · 2026-08-19
The developer of Pangram clarifies that the tool is not a measure of writing quality. A 100% human score does not imply good writing, nor does a 100% AI score imply bad writing. Essentially, Pangram is an authorship attribution algorithm based on near-infinite samples from dozens of LLMs, designed to distinguish between human, AI-assisted, and AI-generated text.
It generalizes to unseen models typically when:
- The LLM is from the same family as a training set model (e.g., Opus 4.5 to 4.6).
- The LLM is trained on synthetic data from a training set model (e.g., Llama SFT'ed on 4o outputs).
- The LLM is similar to a training set model (e.g., trained mostly on common crawl).
More from Models
- Brett Adcock Says Hark Has "Sci-Fi Movie Hardware" in the Lab, AGI Needs More Multimodality — ChrisGPT · 2026-08-19
- DeepSeek's demand destruction strategy hasn't worked yet — zephyr_z9 · 2026-08-19
- Qwen3.8 27B Breaks Size-Intelligence Curve, Resets Consumer Hardware Expectations — smtabatabaie · 2026-08-19
- Analysis Suggests OpenAI's 'Astra' Model Unaffected by RL Slowdown — haider1 · 2026-08-19
- Anthropic's protein design PR dissected: open-source models did most of the lifting — teortaxesTex · 2026-08-19
- Inkling-Small hits Agent Arena: matches Inkling performance at half the price — simonguozirui · 2026-08-19