Dmitry Noranovich

Machine Learning Engineer. I learn by teaching. Opinions are my own.

Building

[Dmitry Noranovich @javaeeeee1]

@freeCodeCamp Discuss the article on Reddit https://t.co/mV5oNStdh3

[Dmitry Noranovich @javaeeeee1]

AI Guardians of the Codebase: How Anthropic, OpenAI, Google, and Others Are Transforming Vulnerability Scanning and Automated Fixes

https://t.co/5HOfQcmy2a

[Dmitry Noranovich @javaeeeee1]

DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression

https://t.co/Ya0KObdejt

[Dmitry Noranovich @javaeeeee1]

Lighting the Way for AI: Photonics in Computing as of Mid-2026

https://t.co/WYlu4Oj10Z

[Sebastian Raschka @rasbt]

Inference scaling part 1.
Starting with a modded text generation function (temperature scaling, top-p filtering, multinomial sampling) to generate diverse outputs for self-consistency and best-of-N (improving answer accuracy by>2x)

00:00 Introduction and recap
00:31 Training-time and inference-time scaling
07:52 What we'll implement
11:47 Notebook setup and model loading
17:43 Building a flexible text generation function
24:40 Chain-of-thought prompting
28:26 Sampling and output diversity
33:43 Next-token logits and greedy decoding
38:20 Temperature scaling step by step
42:46 Softmax and token probabilities
47:42 Multinomial sampling
54:51 Adding temperature sampling to text generation
59:31 Top-p filtering step by step
1:10:23 Adding top-p filtering to text generation
1:13:43 Sampling and LLM watermarking
1:16:01 Self-consistency and majority voting
1:20:36 Implementing self-consistency
1:29:02 MATH-500 results
1:35:01 Accuracy and compute tradeoffs
1:36:50 Next steps and self-refinement

[Qwen @Alibaba_Qwen]

🚀 Meet Qwen3.8-Omni-Flash, Qwen's first omni-modal model built around agentic capabilities!

Native audio-video understanding, reasoning, and tool use come together in one model: understand the content, plan the task, execute with tools, and deliver the result.

Highlights: 🥳
- Audio-video intelligence that gets things done: jointly reason over what's seen and heard, and orchestrate tools across long workflows to auto-edit vlogs, translate short videos, and turn movies into recaps.
- A major leap: approaching Gemini 3.8 Flash in audio-video capabilities; +19.5 points on average in agent performance across WildClawBench-MM & UniClawBench.
- 1M-token context with agentic perception: actively explore long videos and locate key moments with higher accuracy, using 51.8% fewer tokens than static understanding on OmniVideoBench.

Video input costs are reduced by about 89% compared with Qwen3.5-Omni-Plus, making long-form audio-video understanding and agentic workflows more affordable than ever.

To help you build apps around Omni, we're also open-sourcing Qwen-MM-Plugins and Qwen-Live Harness! 🛠️

We can't wait to see what you build with Qwen3.8-Omni-Flash! 👀

- Blog: https://t.co/oM9V1TkqYF
- Qwencloud: https://t.co/Cc2I8ELAnD
- Qwen Studio: https://t.co/V7RmqMaVNZ
- API: https://t.co/lNE7fH5YUt
- Qwen-MM-Plugins: https://t.co/SnM27dDP3d
- Qwen-Live Harness: coming soon
https://t.co/iVYlGjIbdy

[Thariq @trq212]

We're adding support for AGENTS.md to Claude Code.

Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md.

You can toggle this behavior in /config.

[Dmitry Noranovich @javaeeeee1]

@MiniMaxAgent Joint the conversation on Reddit https://t.co/RsklR43vyC

[Connor Shorten @CShorten30]

What happens when one, extremely agent-pilled builder, takes on the insurance brokerage industry?

I am super excited to share the 144th episode of the @weaviatepodcast, featuring Alex Ledbetter (@alexledbetter15) from Delegance Brokerage!

From the product itself to the operations required to run a business, the way Alex embraces Agents is remarkable! 👾

Alex is a self-taught developer (and we dive into how that differs from Vibe Coding 😆), and I learned so much from his perspectives on these tools. From shipping 900 PRs in 3 months with 18 terminal agents running, email and browser agents... It is a very interesting case study in what you can do with these systems.

I also loved diving into his use of Weaviate! Beyond happy to feature case studies like this one on the podcast 💚

YouTube: https://t.co/Xdr9Dq8JUC
Spotify: https://t.co/IsgWZrzc0r

[Dmitry Noranovich @javaeeeee1]

@digitalix Discuss it on Reddit https://t.co/BpmmXcecvw

[Dmitry Noranovich @javaeeeee1]

@AIatAMD Discuss the program on Reddit https://t.co/elHLC2mnAl

[AI at AMD @AIatAMD]

10 things to know before joining the AMD AI Developer Program 👇

1. It’s free. Anyone can join, whether you are a student starting out or a seasoned ML engineer.

2. You get $100 in cloud credits on day one. Enough to prototype, benchmark, and get moving on AMD tech.

3. You don’t need any hardware to start. Build immediately with cloud access.

4. There’s a private Discord with a community of engineers who will provide technical discussion and help.

5. You get early invites to dev events and workshops. Before they’re public. Before they fill up.

6. Monthly Ask the Expert hours with open-source partners. Ask questions. Go deep. Learn the tricks of the trade.

7. Your projects can get featured on official AMD Developer channels and at events.

8. You’re automatically entered into monthly hardware raffles. No forms. No extra steps.

9. You get a free month of @DeepLearningAI Pro. High quality courses for you to expand your skills.

10. The more you participate, the more you unlock. More access, more visibility, more opportunities.

If you’re building, training, or optimizing AI models, we’d love to have you join us!

[Dmitry Noranovich @javaeeeee1]

@genevieve__h Discuss the article on Reddit /r/AgentContext_dev/comments/1wipzm8/ai_ui_design_8_ways_to_make_vibecoded_apps_look/

[Dmitry Noranovich @javaeeeee1]

Appwrite: The open-source cloud for agents and developers by @adityaoberai1, @eldadfux, @DittmannTorsten, and more https://t.co/V0xYw7B9Ev

[Dmitry Noranovich @javaeeeee1]

MCPJam : Test, debug, and evaluate your MCP servers. by @PrathmeshPatel_ https://t.co/ckD5gtUgVj

[Dmitry Noranovich @javaeeeee1]

Amy by Jellyfish: Your AI sourcing employee for recruiting teams by @saymooneh https://t.co/LFzeokjVBZ

[Dmitry Noranovich @javaeeeee1]

Bitrise: The mobile DevOps platform built for the speed of AI by @LisaDziuba https://t.co/0FoHQelnTq

[Dmitry Noranovich @javaeeeee1]

Text Agent Store: A marketplace for AI agents you can text by @stannostudio and @patricksullivan https://t.co/meu9j0T31A

[Dmitry Noranovich @javaeeeee1]

Die With Me: AIM buddy list for your friends’ Claude Code and Codex usage by @bnl5110_ https://t.co/WTUPorwWve

[Dmitry Noranovich @javaeeeee1]

ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments

https://t.co/YXBFRa4IoP