Hackernews posts about Qwen 4B

RSS feed for Qwen 4B

01
newNeuron Explorer for Qwen 4Bgithub.com
▲ 3vpj 13d discuss
02
newShow HN: 4B VLM trained on 600k questions to query images in one passgithub.com
▲ 2aditmag 25d discuss
03
newShow HN: Qwen3-4B with the AI writing tics trained out100toolkit.com
▲ 158erikkannike 19d 💬 113
04
newNeuron Explorer for Qwen3.5-4B: See every neuron at work, make targeted editsgithub.com
▲ 29Schlagbohrer 12d discuss
05
newShow HN: Hob - Open-source, on device alternative to Claude Cowork and ChatGPTgithub.com
▲ 180pancomplex 12d 💬 44
06
newShow HN: Pulse Local, a Mac budgeting app whose AI models run on your machinepulseos.finance
▲ 122niztk 25d 💬 71
07
newTraining Qwen 4B to Beat Large Models on Work Tasksneurometric.substack.com
▲ 16robmay 268d discuss
08
newShow HN: Dante-Qwen-4B – Curing LLM "Neurosis" with a Divine Comedy Curriculumhuggingface.co
▲ 5hunterbown 344d 💬 1
09
newGood Results when training Qwen 3 4B to learn a new domainwww.teachmecoolstuff.com
▲ 5dev-experiments 78d discuss
10
newI've tested some local LLMs on prosumer hardware, here are some findings
▲ 7felineflock 76d discuss
11
newShow HN: Distilled 0.6B text-to-SQL modelgithub.com
▲ 5maciejgryka 290d discuss
12
newShow HN: An offline first focused agentic CLI application powered by Ollama
▲ 4g023 227d discuss
13
newSumi – Open-source voice-to-text with local AI polishing
▲ 2alkd 244d discuss
14
newShow HN: Promptscout a local prompt enricher for Claude Codegithub.com
▲ 2obsfx 268d discuss
15
newQwen3-4B-Thinking-2507huggingface.co
▲ 198IdealeZahlen 458d 💬 61
16
newOn-device Qwen benchmark: the 4B ties the 9B, and more findingsgithub.com
▲ 2thomasblanc 116d discuss
17
newShow HN: Building a full agentic harness around a 4B model is hardorvena.app
▲ 5mehrant 80d discuss
18
newQwen 3.5 9B, 4B models beating 30B, 80B modelshuggingface.co
▲ 2satvikpendem 250d discuss
19
newQwen 3.5: 9B, 4B, 2B, 0.8Bhuggingface.co
▲ 7tosh 250d discuss
20
newRunning Qwen 3.6 35B-A3B-4B on MacBook Pro M5 64GB with toolswww.youtube.com
▲ 4elvean 202d 💬 1
21
newShow HN: I fine-tuned Qwen 3.5 (0.8B–4B) on a Mac for text-to-SQL – 2B beats 12Bgithub.com
▲ 7sciences44 247d 💬 1
22
newShow HN: Jan-nano, 4B agentic model that outperforms DeepSeek-v3-671B using MCPtwitter.com
▲ 6emreckartal 509d discuss
23
newIs anyone compressing AI models for the 4B people without GPUs or internet?
▲ 3yashpxl 251d 💬 1
24
newFine-tuning Qwen3.5-4B to replace Gemini Flash-Lite in a RAG pipelineduane.sh
▲ 1duanemb 39d discuss
25
new4B parameter Deep Research model based on Qwenhuggingface.co
▲ 1sumo43 376d discuss
26
newLaunch HN: RunRL (YC X25) – Reinforcement learning as a servicerunrl.com
▲ 71ag8 416d 💬 22
27
newShow HN: KARMA – An evaluation framework for Medical AI systemskarma.eka.care
▲ 2k2so 453d discuss
28
newShow HN: Jiwo - Small decision models topping decision index leaderboardgithub.com
▲ 1jiwidi 33d discuss
29
newQwen3-4B-Thinking: "This is art–pelicans don't ride bikes "simonwillison.net
▲ 4nnx 451d discuss
30
newI swept all 667 layer duplication configs for Qwen3-4B on 1x RTX 3090austinsnerdythings.com
▲ 3auspiv 206d discuss
31
newQwen3-VL-4B and 8B runs locally on NPU, GPU, and CPU with one SDKnexa.ai
▲ 3alanzhuly 389d discuss
32
newTeaching Qwen3-4B to Trade: From Hold-Collapse to +9.4% Returnssabareesh.com
▲ 2sabareesh 239d discuss
33
newLatest progress helping Qwen3-4B Learngithub.com
▲ 2kibbyd1985 252d discuss
34
newFine‑tuning Qwen3‑4B on a MacBook M1 for customer supportblog.layerx.xyz
▲ 1felltrifortence 204d discuss
35
newA fine-tune of Qwen3-4B for use with Cavil for legal text classificationhuggingface.co
▲ 1Tomte 501d discuss
36
newLaunch HN: RunAnywhere (YC W26) – Faster AI Inference on Apple Silicongithub.com
▲ 240sanchitmonga22 242d 💬 153