ResearchBuzz: FirehoseThe Register: GitHub ponders kill switch for pull requests to stop AI slop. “AI slop has come home to roost and GitHub wants help from its community of software developers to figure out how to manage the mess.”
https://rbfirehose.com/2026/02/06/the-register-github-ponders-kill-switch-for-pull-requests-to-stop-ai-slop/
Associated Press: Hank and John Green’s studio becomes a nonprofit as they aim to make ‘trustworthy content’ online. “The change is intended to ensure viewers have access to engaging, fact-based content that can compete free of advertisers’ interests in the attention economy. It comes as artificial intelligence gives rise to absurdist ‘brain rot’ and distorted deepfake images while […]
https://rbfirehose.com/2026/02/06/associated-press-hank-and-john-greens-studio-becomes-a-nonprofit-as-they-aim-to-make-trustworthy-content-online/
Associated Press: Hank and John Green’s studio becomes a nonprofit as they aim to make ‘trustworthy content’ online
Associated Press: Hank and John Green’s studio becomes a nonprofit as they aim to make ‘trustworthy content’ online. “The change is intended to ensure viewers have access to engaging, fact-ba…
Reddit Tech VN BotCông cụ AntiSlop giúp phát hiện các vấn đề trong mã code được tạo bởi AI như hàm.stub, câu lệnh điều khiển, bình luận tạm thời và lỗi không được xử lý. Hỗ trợ Rust, Python, JavaScript/TypeScript, Go. #AntiSlop #AICode #Linter #CodeReview #MãHóa #PhátTriểnPhầnMềm #CôngCụPhátTriển
Công cụ AntiSlop giúp phát hiện mã AI thiếu an toàn: hàm bỏ trống, `console.log`, bình luận "tạm thời", lỗi chưa xử lý – dùng tree-sitter để phân tích chính xác, không báo nhầm trong chuỗi. Hỗ trợ Rust, Python, JS/TS, Go. Dùng CLI, cài qua cargo/npm. Hữu ích cho team dùng AI viết code. #AI #CodeQuality #DeveloperTools #AntiSlop #DevOps #LậpTrình #ChấtLượngMãNguồn #CôngCụLậpTrình
Jake in the desertAustralian artist and environmental engineer Tega Brain has created 'Slop Evader', a browser extension search tool for either Firefox or Chrome that evades AI slop by using the Google search API to only return content created before ChatGPT's first public release on November 30, 2022: https://tegabrain.com/Slop-Evader
#AISlop #AntiSlop #AntiAI #SlopEvader #search #slop #Chrome #Firefox #Google #SearchTools #BrowserExtensions
N-gated Hacker News🎉🎉 Presenting "Antislop": the magical 🪄 framework designed to obliterate repetitive patterns in language models, because redundancy is SO last year! 🤦♂️ Now, you too can donate for the cause and keep the world safe from the horrors of linguistic déjà vu. 🙃
https://arxiv.org/abs/2510.15061 #Antislop #LanguageModels #Innovation #TechForGood #RedundancyFree #HackerNews #ngated
https://arxiv.org/abs/2510.15061 #Antislop #LanguageModels #Innovation #TechForGood #RedundancyFree #HackerNews #ngated
Antislop: A Comprehensive Framework for Identifying and Eliminating Repetitive Patterns in Language Models
Widespread LLM adoption has introduced characteristic repetitive phraseology, termed ``slop,'' which degrades output quality and makes AI-generated text immediately recognizable. We present Antislop, a comprehensive framework providing tools to both detect and eliminate these overused patterns. Our approach combines three innovations: (1) The Antislop Sampler, which uses backtracking to suppress unwanted strings at inference time without destroying vocabulary; (2) An automated pipeline that profiles model-specific slop against human baselines and generates training data; (3) Final Token Preference Optimization (FTPO), a novel fine-tuning method that operates on individual tokens, surgically adjusting logits wherever a banned pattern has appeared in an inference trace. We demonstrate that some slop patterns appear over 1,000$\times$ more frequently in LLM output than human text. The Antislop Sampler successfully suppresses 8,000+ patterns while maintaining quality, whereas token banning becomes unusable at just 2,000. Most importantly, FTPO achieves 90\% slop reduction while maintaining or improving performance in cross-domain evals including GSM8K, MMLU, and creative writing tasks. In contrast, DPO suffers significant degradation in writing quality and lexical diversity despite achieving weaker suppression. We release all code and results under MIT license: https://github.com/sam-paech/auto-antislop.
Hacker NewsAntislop: A Framework for Eliminating Repetitive Patterns in Language Models
https://arxiv.org/abs/2510.15061
#HackerNews #Antislop #LanguageModels #AI #Innovation #RepetitivePatterns #Framework
Antislop: A Comprehensive Framework for Identifying and Eliminating Repetitive Patterns in Language Models
Widespread LLM adoption has introduced characteristic repetitive phraseology, termed ``slop,'' which degrades output quality and makes AI-generated text immediately recognizable. We present Antislop, a comprehensive framework providing tools to both detect and eliminate these overused patterns. Our approach combines three innovations: (1) The Antislop Sampler, which uses backtracking to suppress unwanted strings at inference time without destroying vocabulary; (2) An automated pipeline that profiles model-specific slop against human baselines and generates training data; (3) Final Token Preference Optimization (FTPO), a novel fine-tuning method that operates on individual tokens, surgically adjusting logits wherever a banned pattern has appeared in an inference trace. We demonstrate that some slop patterns appear over 1,000$\times$ more frequently in LLM output than human text. The Antislop Sampler successfully suppresses 8,000+ patterns while maintaining quality, whereas token banning becomes unusable at just 2,000. Most importantly, FTPO achieves 90\% slop reduction while maintaining or improving performance in cross-domain evals including GSM8K, MMLU, and creative writing tasks. In contrast, DPO suffers significant degradation in writing quality and lexical diversity despite achieving weaker suppression. We release all code and results under MIT license: https://github.com/sam-paech/auto-antislop.
gabeWorking on a Django middleware and template tags when I return from vacation, that will randomly inject some nice ai slop into the return when they meet one of those user agents.
Maybe have it do a 301 to one of the speedtest files at random.
Or never close the connection, just keep it open indefinitely for the max timeout so they cant hammer your biz.
Probably configurable in your settings dot py
#django #antislop
Maybe have it do a 301 to one of the speedtest files at random.
Or never close the connection, just keep it open indefinitely for the max timeout so they cant hammer your biz.
Probably configurable in your settings dot py
#django #antislop