Control your X/Twitter feed using a small on-device LLM
We built a Chrome extension and iOS app that filters Twitter's feed using Qwen3.5-4B for contextual matching. You describe what you don't want in plain language—it removes posts that match semantically, not by keyword. What surprised us was that because Twitter's ranking algorithm adapts based on what you engage with, consistent filtering starts reshaping the recommendations over time. You're implicitly signaling preferences to the algorithm. For some of us it "healed" our feed. Currently running inference from our own servers with an experimental on-device option, and we're working on fully…
Control your X/Twitter feed using a small on-device LLM launched on April 9, 2026 with 15 votes, #1,070 of 21,248 launches that month and more than 90% of that year's launches.
In plain words
A Chrome extension and iOS app that filters Twitter posts using semantic matching rather than keywords. Users describe unwanted content in plain language, and an on-device language model removes semantically similar posts from their feed. Because Twitter's algorithm adapts to engagement patterns, consistent filtering gradually reshapes feed recommendations. The tool runs inference on company servers with experimental local processing in development. It handles figurative language imperfectly but outperforms keyword-based muting. No user data is collected.
written from the facts on this page · September 2026
From the sources
Describe what to filter from your feed, and Bouncer removes it. My Twitter feed was a cesspool because it was beyond my control. Its algorithm, optimized for addiction rather than enjoyment, would discover how best to push my buttons. Every day I was subjected to tweets that made me feel left out, angry, and defeated. But I couldn’t stop scrolling. A super-intelligent algorithm controlled my attention and fed me poison. Eventually I stopped using the “For You” tab altogether, and I felt happier than I had in a long time. Despite its flaws, the algorithm had guided me to some interesting events where I made real-life friends. No platform is better for real-time information, and no platform…from imbue.com
In the maker’s words, at launch
We built a Chrome extension and iOS app that filters Twitter's feed using Qwen3.5-4B for contextual matching. You describe what you don't want in plain language—it removes posts that match semantically, not by keyword. What surprised us was that because Twitter's ranking algorithm adapts based on what you engage…
Does a similar job
all alternatives →
I built a self hosted recommendation feed to escape Google's algorithm2022 · github.com · ▲214I created this chrome extension for myself where I track my own behavior locally and recommend myself content from platforms I want content from (youtube/twitter/quora/etc) in a feed. I made it public just in case anyone else was interested. I would rather have control over my own algorithm and own the data. Also, it gives me flexibility. Turns out I do like these feeds just not when I don't own it haha. Let me know what you think of my implementation?
Filter out engagement bait and politics on your X/Twitter feed2025 · unbaited.danielpetho.com · ▲71hi friends! i'm pretty tired of engagement bait and all the political nonsense on my x/twitter feed. i was curious if i could use an llm to filter out these type of content, so i prototyped a quick chrome extension. it uses LLama 3.3 to analyze the tweet through https://groq.com/ (because they are super-super fast). the extension is available in the chrome store, also there is a link to the repo. - you can tweak the system prompt for the filtering - but you need your own API key from Groq (you can get one for free)
FilterIn2025 · chromewebstore.google.com · ▲148Hide LinkedIn posts containing keywords you want to avoid
More ai this month
the category →
Paintings by AI models: programs run through a simulation of oil paint.
AI · 20d ago · stillwet.art
PixVerse R2▲381A real-time world model you can explore and change
AI · 29d ago · world.pixverse.video
Source: Show HN post ↗ · launch text quoted in part
Launched alongside, April 2026
the whole month →
- AG
Thought the resources for GPU arch were lacking, so here we are
Life & fun · Apr 2026 · jaso1024.com
Built a ~9M param LLM from scratch to understand how they actually work. Vanilla transformer, 60K synthetic conversations, ~130 lines of PyTorch. Trains in 5 min on a free Colab T4. The fish thinks the meaning of life is food. Fork it and swap the personality for your own character.
AI · Apr 2026 · github.com
I made a laptop stand out of solid concrete, in the style of brutalist architecture, with themes of urbex and urban decay
Life & fun · Apr 2026 · sam-burns.com
With social media and now AI, its important to keep the indie web alive. There are many people who write frequently. Blogosphere tries to highlight them by fetching the recent posts from personal blogs across many categories. There are two versions: Minimal (HN-inspired, fast, static): https://text.blogosphere.app/ Non-minimal: https://blogosphere.app/ If you don't find your blog (or your favorite ones), please add them. I will review and approve it.
AI · Apr 2026 · text.blogosphere.app