ZK
Editor-in-Chief · CPO of AlpinaGPT

Zhemal Khamidun

Head of AI at Alpina Digital, CPO of AlpinaGPT (~30-person engineering & AI team). Founder of Vibe Academy & Hamidun Consulting, author of the AI channel «Готовим ИИшницу» (@jemal_hamidun).

10,426Published articles
50+Verified primary sources
6Language editions (RU/EN/ES/PT/FR/AR)
DeepSeek open-sources its DeepInfer inference engine, and analysts calculate an 82% margin
Business·Jiqizhixin (机器之心)

DeepSeek open-sources its DeepInfer inference engine, and analysts calculate an 82% margin

Oct 22, 2026·3 min
RVCBench Tests 18 Voice Cloning Models on Noise, Compression, and Anti-Cloning Protection
Research

RVCBench Tests 18 Voice Cloning Models on Noise, Compression, and Anti-Cloning Protection

Oct 22, 2026·2 min
The Expectations Gap: 91% of Anthropic Employees and 39% of Americans Expect a Nobel-Level AI Breakthrough
LLM

The Expectations Gap: 91% of Anthropic Employees and 39% of Americans Expect a Nobel-Level AI Breakthrough

Oct 22, 2026·2 min
Anthropic: AI Agents Now Outnumber Employees 6 to 1, and Junior Hiring Is Shrinking
Business

Anthropic: AI Agents Now Outnumber Employees 6 to 1, and Junior Hiring Is Shrinking

Oct 22, 2026·2 min
LILIE teaches image generation through language, cutting gFD from 66.84 to 53.01
LLM·Hugging Face Daily Papers

LILIE teaches image generation through language, cutting gFD from 66.84 to 53.01

LILIE's reversible encoder gives a generative model a language signal instead of a purely pixel-based loss: gFD on GPIC drops by about 14 points, and code and…

Oct 22, 2026·2 min
Competence Over Accuracy: Why GPT and Qwen Lose Logic When Problems Get Rephrased
LLM·Hugging Face Daily Papers

Competence Over Accuracy: Why GPT and Qwen Lose Logic When Problems Get Rephrased

A new methodology called Competence-Based Evaluation (CBE) has uncovered a fundamental flaw in LLM logical reasoning: even strong GPT and Qwen models fail on the…

Oct 22, 2026·2 min
A no-hair theorem for transformers: the prompt is forgotten as t^-α, α≈0.55–0.7
Research·Hugging Face Daily Papers

A no-hair theorem for transformers: the prompt is forgotten as t^-α, α≈0.55–0.7

After t tokens, the distance between hidden states of different prompts decays as t^-α. Trained models show α≈0.55–0.7; the proven bound is O(P(log t)^{L-1}/t).

Oct 22, 2026·2 min
Huginn Beats Ouro on All 10 Benchmarks: The Design Space of Recurrent LLMs
Research·Hugging Face Daily Papers

Huginn Beats Ouro on All 10 Benchmarks: The Design Space of Recurrent LLMs

Recurrent language models reuse Transformer blocks, but getting the architecture right is a scarce skill. Researchers at the Institute of Foundation Models…

Oct 22, 2026·2 min
Top themes
01MOSS-Image: Open 5B Model Renders Long Text with 96.3% CoverageOct 22, 2026·2 min02MIST shows video world models don't forget; they fail to carry out the commandOct 22, 2026·2 min03"Wait" in Reasoning Models Turns Out to Be Inherited from Data, Not Discovered by RLOct 22, 2026·3 min04Vercel Shop Reaches GA: Shopify Storefronts Learn to Sell to AI AgentsOct 22, 2026·3 min05OpenAI's Interpretability Founder Says 90% of Neurons Have a Single FunctionOct 22, 2026·2 min
Kunlun Wanwei launches Matrix-Game 4.0: a video model that plays in 480p at 60 FPS on a single GPU
Hardware·Jiqizhixin (机器之心)

Kunlun Wanwei launches Matrix-Game 4.0: a video model that plays in 480p at 60 FPS on a single GPU

Oct 22, 2026·2 min
DeepSeek releases V4.1-Flash-1017 and opens its Search API at $0.75 per 1000 calls
LLM·DeepSeek Blog

DeepSeek releases V4.1-Flash-1017 and opens its Search API at $0.75 per 1000 calls

Oct 22, 2026·2 min
Claude Traces the Origin of Five NIST Curve Seeds in Half an Hour
Research·Jiqizhixin (机器之心)

Claude Traces the Origin of Five NIST Curve Seeds in Half an Hour

Oct 22, 2026·2 min
HumanoidPose Estimates Humanoid Robot Pose From Video, Cutting Joint-Angle Error by 18–32%
Robotics·Hugging Face Daily Papers

HumanoidPose Estimates Humanoid Robot Pose From Video, Cutting Joint-Angle Error by 18–32%

Oct 22, 2026·2 min
Researchers Annotate 2,122 WildChat Conversations: 71.1% of Legal Chats Seek Personal Advice
Business·Hugging Face Daily Papers

Researchers Annotate 2,122 WildChat Conversations: 71.1% of Legal Chats Seek Personal Advice

A taxonomy of legal queries to LLMs, 2122 WildChat-1M conversations and open models: 71.1% of chats with legal content ask for personalized or practical advice.

Oct 22, 2026·2 min
StopLoss Cuts Off Doomed Coding-Agent Runs and Saves Up to 48% of Costs
Business·Hugging Face Daily Papers

StopLoss Cuts Off Doomed Coding-Agent Runs and Saves Up to 48% of Costs

StopLoss, a small external monitor, predicts agent failure from the latest steps and halts the run: 4.5% lower cost on Verified and up to 31% on SWE-bench Pro.

Oct 22, 2026·2 min
Changing the definition of "unsafe" moves a guardrail score from 0 to 100 with 0.03% of weights
Research·Hugging Face Daily Papers

Changing the definition of "unsafe" moves a guardrail score from 0 to 100 with 0.03% of weights

The authors labeled ~53 thousand prompts twice under different policies and trained 20 guardrail models. Benchmark spread reached 23x, and some test sets were already in the training data.

Oct 22, 2026·3 min
Anthropic to put 5% of its shares into a Public Benefit Fund, with the stake growing as the company grows
Business·Anthropic Blog

Anthropic to put 5% of its shares into a Public Benefit Fund, with the stake growing as the company grows

Oct 22, 2026·2 min
Researchers Propose Anti-Persistence: LLM Agent Memory Without Stored Rules
Research·Hugging Face Daily Papers

Researchers Propose Anti-Persistence: LLM Agent Memory Without Stored Rules

Oct 22, 2026·2 min
3D Scene Generation: DiT Works in a 3D Latent and Encodes a Scene in 1K Tokens
Research·Hugging Face Daily Papers

3D Scene Generation: DiT Works in a 3D Latent and Encodes a Scene in 1K Tokens

Oct 22, 2026·2 min
POPE audit: the choice of "yes/no" tokens shifts LLaVA-1.5 F1 by 6.13 points
Research·Hugging Face Daily Papers

POPE audit: the choice of "yes/no" tokens shifts LLaVA-1.5 F1 by 6.13 points

Oct 22, 2026·2 min
Parallel Refinement Instead of Diffusion: DRLMs Hit 93.9% on Sudoku-Extreme
Research·Hugging Face Daily Papers

Parallel Refinement Instead of Diffusion: DRLMs Hit 93.9% on Sudoku-Extreme

Oct 22, 2026·2 min
TokenWorth Teaches AI Agents to Forget: Best Results on Three Environments With the Fewest Tokens
Research

TokenWorth Teaches AI Agents to Forget: Best Results on Three Environments With the Fewest Tokens

Oct 22, 2026·2 min
Ropedia W0 trained on 10,000 hours of human video and 3,000 hours of robot data
Robotics

Ropedia W0 trained on 10,000 hours of human video and 3,000 hours of robot data

Oct 22, 2026·2 min
Anthropic: junior share of hires fell from 27% to 9%, while agent hours outpaced human hours 6x
Business

Anthropic: junior share of hires fell from 27% to 9%, while agent hours outpaced human hours 6x

Oct 22, 2026·3 min
TokenGate speeds up LLM prefill by up to 11.3x with learnable token pruning
Business·Hugging Face Daily Papers

TokenGate speeds up LLM prefill by up to 11.3x with learnable token pruning

Liana Patel, Rishabh Ranjan, Ion Stoica, Matei Zaharia, and Carlos Guestrin of UC Berkeley have introduced TokenGate to accelerate prefill in LLMs. The mechanism…

Oct 22, 2026·2 min
Vercel sets deadlines for vendors: static keys must go by 2027
Business·Vercel Blog

Vercel sets deadlines for vendors: static keys must go by 2027

Vercel surveyed 1,000+ SaaS vendors: only 8.4% support secure authentication. From November 30, its marketplace will flag "Not secure" listings; from January 2027…

Oct 22, 2026·3 min