Jul 17, 2026

Kimi K3: Open Frontier Intelligence

Kimi K3 is a groundbreaking 2.8 trillion-parameter model that integrates advanced Kimi Delta Attention and Attention Residuals, enabling it to excel in long-horizon coding and knowledge work while being the first open model in the 3T-class category.

Lingbot-map: A 3D foundation model for reconstructing scenes from streaming data

LingBot-Map introduces a Geometric Context Transformer that integrates coordinate grounding, dense geometric cues, and long-range drift correction, achieving stable streaming 3D reconstruction at approximately 20 FPS over sequences exceeding 10,000 frames.

Mozilla: The state of open source AI

Open-source AI is now a dominant force, with open-weight models accounting for over 50% of production tokens by mid-2026, reflecting a significant shift in developer preference towards cost-effective and accessible solutions.

Detecting LLM-Generated Texts with “Classical” Machine Learning

Mainstream LLM-generated text can be distinguished from human-written content with approximately 85% accuracy using traditional machine learning models like Linear SVC and Naive Bayes. This detection leverages statistical patterns inherent in AI-generated text, which are often consistent and detectable compared to the variability of human writing.

Kimi K3, and what we can still learn from the pelican benchmark

Kimi K3, with 2.8 trillion parameters, is hailed as the first "open 3T-class model," outperforming many competitors in benchmarks while promising an open weight release by July 27, 2026.

Mathematics of Data Science

This book delves into the mathematical foundations of data science, covering essential topics such as Singular Value Decomposition, Linear Regression, and Deep Learning. Link to article

Ring-Zero: Scaling Zero RL to a Trillion Parameters for Emergent Reasoning

Scaling to 1 trillion parameters in zero RL significantly boosts sample efficiency and performance, revealing advanced cognitive behaviors like self-verification and parallel reasoning that eliminate the need for hand-crafted heuristics.

Pseudpocalypse

Pseudpocalypse suggests that unique writing styles can be statistically linked to authors, potentially compromising pseudonymity as machine learning advances in text analysis.

AI Meets Cryptography 2: What AI Found in OpenVM's ZkVM

AI auditor zkao identified a critical soundness bug in OpenVM's openvm-pairing library, allowing malicious actors to forge pairing equality, which undermines the integrity of cryptographic proofs built on this library.

PennyLane is an open-source quantum software platform for quantum

PennyLane is an open-source quantum software platform that facilitates the creation of quantum algorithms across various domains, including quantum computing, quantum machine learning, and quantum chemistry.

Simulating everything, sort of: The promise and limits of world models

World models are emerging as a promising alternative to large language models (LLMs), focusing on simulating the physical world for applications in robotics, research, and asset generation, with significant funding and advancements in recent months.

Our approach to bioresilience

Google DeepMind's initiative, AlphaGenome, leverages AI to enhance our understanding of the genome, aiming to unlock new avenues in bioresilience and health innovation. Learn more.

🤗NVIDIA Nemotron 3 Embed Ranks #1 Overall on RTEB, Advancing Agentic Retrieval

NVIDIA's Nemotron 3 Embed has achieved the top rank on the RTEB leaderboard, showcasing significant advancements in agentic retrieval capabilities.

ExTernD: Expanded-Rank Ternary Decomposition Ternary LLM PTQ with Accuracy Approaching Any Quantization Level [P]

ExTernD introduces a novel approach to ternary post-training quantization (PTQ) by decomposing matrices into two ternary matrices and an inner diagonal scaling matrix, allowing for arbitrary accuracy adjustments without fixed matrix sizes.

Show HN: ReasonGate- An explainable gate that blocks LLM prompt injection

ReasonGate is an explainable security gate for LLM applications that audits every decision, effectively blocking attacks like indirect injection before they reach the model, as demonstrated in a real-world banking scenario where it prevented a $84,200 breach.