← All sources

PyTorchKR news

The latest AI and developer news from PyTorchKR, with short summaries.

Visit website →

NEWvLLM Introduces Hardware-Agnostic Layer Architecture to Support Diverse Accelerators

·1·2026.09.25 19:00

NEWvLLM rope_scaling Override Causes 36% Model Configuration Mismatch

·2026.09.25 17:00

NEWvLLM Releases 'Decision 1.0', AI Models Specialized for Decision-Making

·2026.09.25 10:00

PickAnthropic Releases Migration Guide for Claude Opus 5.5

·2026.09.24 18:00

PickTrillion Labs Releases GravityOCR, an OCR Model Applying Self-Speculative Decoding

·2026.09.24 11:00

NVIDIA Redefines AI Agent Evaluation Metrics

·2026.09.23 17:00

StepFun Releases Step 5 Preview

·2026.09.23 12:00

PickPostgreSQL Extension pgjev Released, Supporting Natural Language Filtering Without Embeddings

·2026.09.23 07:00

NVIDIA Releases AIPerf, an LLM Inference Benchmarking Tool

·2026.09.22 18:00

Google Research Unveils R4T, an RL-Based Search Distillation Technique

·2026.09.22 15:00

Bend 2 Released: A Proof-Oriented Language That Fails Compilation on AI Code Rule Violations

·2026.09.22 07:00

Context Compression Plugin 'Fast Jev Compaction' for Claude Code Released

·2026.09.21 16:00

OpenArm Releases Open-Source Dual-Arm Robot and Evaluation Platform

·1·2026.09.21 09:00

Inco AI Releases Splash, an Inference Engine Dedicated to Apple Silicon

·2026.09.20 11:00

UC Berkeley Publishes 'Harness Tax' Study: Up to 5x Cost Difference Between Harnesses

·2026.09.19 18:00

PrismML Releases 'Ternary Bonsai 2 27B', a Ternary Quantized Model Maintaining 98.2% of FP16 Performance

·2026.09.19 11:00

Transformer Performance: Mechanism Diversity Is Key Over Placement

·1·2026.09.18 12:00

RLT: Proposed Recurrent Decoder Structure with Computational Depth Increasing with Token Length

·2026.09.18 12:00

Recipe Released for Serving 320B GLM-5.3-Flash on Free Kaggle TPUs

·2026.09.18 07:00

BrowserSkill Released: Agent CLI That Reuses Logged-In Browser Sessions

·2026.09.17 17:00