Collabora presents "Bringing BitNet to ExecuTorch via Vulkan" at PyTorch Conference Europe in Paris (April 7-8) and attends ICLR in Rio de Janeiro (April 23-27). Connect with our team to discuss machine learning and open source innovation!
@pytorch #PyTorchCon #PyTorch2026 #PyTorch #PyTorchFoundation #MachineLearning #DeepLearning #ICLR2026
AI 메모리 병목을 3비트로 해결, 구글 TurboQuant 8배 속도 달성한 방법
구글 리서치가 발표한 TurboQuant는 LLM의 KV 캐시를 3.5비트로 압축하면서 정확도 손실 없이 최대 8배 빠른 처리 속도를 달성한 벡터 양자화 알고리즘입니다.
TurboQuant에 “Pied Piper”라는 별명이 붙은 이유, 그리고 그 비교가 과장인 이유
구글 TurboQuant 발표에 업계가 'Pied Piper', 'DeepSeek 모멘트'라 반응한 이유와 그 비교가 과장인 이유를 분석합니다.
Google's TurboQuant Compresses AI Memory by 6x — With Zero Accuracy Loss
https://techlife.blog/posts/google-turboquant
#Google #TurboQuant #LLM #AIEfficiency #KVCache #ICLR2026 #MachineLearning #Compression
Google Research published TurboQuant, a training-free compression algorithm that shrinks LLM key-value cache memory by at least 6x and speeds up attention by up to 8x on H100 GPUs — without any accuracy penalty.
In a new #ICLR2026 publication we provide a novel algorithm for semi-analytically constructing the stable and unstable manifolds of fixed points and cycles of ReLU-based RNNs:
https://openreview.net/pdf?id=EAwLAwHvhk
Why is this important?
Because it provides insight into why and how trained RNNs produce their behavior, as important for scientific and medical applications and explainable AI more generally. In scientific ML, RNNs are a common tool for *dynamical systems reconstruction* (https://www.nature.com/articles/s41583-023-00740-7), where models are trained to approximate the dynamical system underlying observed time series. Trained RNNs are then to be analyzed further as formal surrogates of the systems trained on.
An RNN’s dynamical repertoire depends on the topological and geometrical properties of its state space. Stable and unstable manifolds of fixed and periodic points dissect a dynamical system’s state space into different *basins of attraction*, their intersections lead to chaotic dynamics with fractal geometry, and – more generally – they provide a type of skeleton for the system’s dynamics, forming structures like separatrix cycles or heteroclinic channels.
Big congratulations to all authors! 🚀
#ICLR2026 #MachineLearning #AIResearch #RepresentationLearning #InformationRetrieval #DenseRetrieval #SelfSupervisedLearning #LanguageModels #NLP #UKPLab #ICLR
📢 𝗨𝗞𝗣 𝗟𝗮𝗯 𝗮𝘁 𝗜𝗖𝗟𝗥𝟮𝟬𝟮𝟲 📢
Happy to share that our paper has been accepted to #ICLR2026 🎉
📜 𝚁𝚎𝚟𝚎𝚕𝚊: 𝗗𝗲𝗻𝘀𝗲 𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗲𝗿 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝘃𝗶𝗮 𝗟𝗮𝗻𝗴𝘂𝗮𝗴𝗲 𝗠𝗼𝗱𝗲𝗹𝗶𝗻𝗴
👥 Fengyu Cai, Tong Chen, Xinran Zhao, Sihao Chen, Hongming Zhang, Sherry Tongshuang Wu, Iryna Gurevych, Heinz Koeppl
ICLR 2026 tổng hợp: Cộng đồng nghiên cứu tập trung vào GRPO (157 bài) thay vì DPO, ưu tiên RLVR (125 bài) thay vì RLHF, và 202 bài về Mamba/SSMs. Nait (tuning thông minh chỉ 10% dữ liệu) giúp tối ưu hiệu quả. 257 bài về tính toán lúc test, 123 bài về hallucination. Cảnh báo: mô hình tuân thủ tốt dễ bị tấn công injection. #AI #HọcMáy #ICLR2026 #NCKH #DeepLearning #Mamba #RLVR #GRPO #MạngNeural #BảoMậtAI #ViễnTưởngAI
https://www.reddit.com/r/LocalLLaMA/comments/1qsh7dz/analyzed_5357_iclr_2026_acc
Hệ thống đa tác nhân thất bại vì 5 vấn đề chính: độ trễ, chi phí token, sai sót dây chuyền, cấu trúc giòn và khả năng quan sát. 14 bài báo ICLR 2026 đề xuất các giải pháp như hành động dự đoán, chia sẻ KV và cây quyết định có cấu trúc. #HệThốngĐaTácNhân #ICLR2026 #TríTuệNhânTạo #MultiAgentSystems #AI
https://www.reddit.com/r/LocalLLaMA/comments/1qs5t82/14_iclr_2026_papers_on_why_multiagent_systems/
Language Technology Group at the University of Oslo has two papers accepted at #ICLR2026!
- Dual Language Models: Balancing Training Efficiency and Overfitting Resilience by David Samuel and Lucas Charpentier
- Fluent Alignment with Disfluent Judges: Post-training for Lower-resource Languages by David Samuel, Lilja Øvrelid, Erik Velldal and Andrey Kutuzov
Details and links in the thread:
Collabora
AI Sparkup
TechLİfe
DurstewitzLab
UKP Lab
Reddit Tech VN Bot
LTG Oslo