arXiv
LACE: Layer-Wise Compression for Dynamic Frame Rate Codecs
Neural audio codecs are a key component in speech language modeling.
HABERLER
Tüm Haberler
Yapay zekâ dünyasından son haberler, derinlikli analizler, teknik incelemeler.
23Yazı
2Yazar
1.8KOkuma · 30g
60Bu Hafta
EDİTÖR seçimi
RSS · BU KATEGORİ →
xAI, Imagine API’yi 2.0’a Yükseltmeye Hazırlanıyor: Görüntü ve Video Tek API’de Birleşiyor
Yapay Zeka HaberleriMicrosoft MAI-Cyber-1-Flash’ı Duyurdu
Microsoft, İlk Siber Güvenlik Modeli MAI-Cyber-1-Flash’ı ve MDASH Sistemini Duyurdu Microsoft, 27 Temmuz 2026 tarihinde siber güvenlik odaklı ilk kendi modelini ve…
Yapay Zeka HaberleriMoonshot AI, Kimi K3 Model Ağırlıklarını ve Teknik Raporunu Açık Kaynak Olarak Yayınladı
Moonshot AI, 27 Temmuz 2026 tarihinde Kimi K3 modelinin tam model ağırlıklarını (weights) ve teknik raporunu resmi olarak açık kaynak olarak yayınladı.…
arXiv
ENCP: Episode-Normalized Conformal Prediction for Vision-and-Language Navigation
Uncertainty estimation for Vision-Language-Navigation (VLN) models is a critical task since it can h...
arXiv
Verifiable Social Reasoning for LLM Assistants
LLM assistants are widely used for daily social advice, yet evaluating their social reasoning in suc...
arXiv
LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence
We introduce LimiX-2, a new model in the LimiX family, developed through model and data scaling guid...
arXiv
Det-LIME: Detector-Aware, Multi-Instance Local Interpretable Model-Agnostic Explanations for Automated Marine Mammal Detection
Despite the rapid uptake of black-box object detectors in marine mammal research and monitoring, exp...
arXiv
JustFit: 200K-Token LLM Serving on a 24 GiB Laptop with Just-in-Time State Management
Capable open-weight models make local coding and reasoning attractive, but their context and executi...
arXiv
Coupled Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback
Large language model (LLM) distillation aims to transfer the capabilities of a powerful teacher to a...
arXiv
Decomposition Buys Integrity, Not Yield
Multi-agent systems split a task across a tree of agents and justify the split with folklore: smalle...
arXiv
Agentic Societies Need a Social Harness
An agentic society is a collection of AI agents that coordinate autonomously across trust boundaries...
arXiv
ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents
We introduce and release ScienceBuddy, an interactive scientific research workspace that brings cont...
arXiv
PhysStream: Streaming Physics-Grounded Video Generation with Structured Scene Memory and Fine-Grained Motion Control
Interactive control for video generation is moving from coarse prompts toward fine-grained, physical...
arXiv
When Should LLMs Abstain? Chain-of-Self-Questioning for Selective Risk Control
Large language models can produce fluent answers when their factual support is weak.
arXiv
Stellar Colosseum: A Many-Agent Harness for Long-Horizon Research in Mathematics and Theoretical Computer Science
Language models can produce plausible short proofs, but may still be unreliable on long-horizon rese...
arXiv
Doğallığın Ötesinde: Otomatik Metin-Konuşma Değerlendiricilerini Dilbilimsel Temelli Boyutlar Üzerinde İncelemek
Otomatik Metin-Konuşma (TTS) değerlendirme yöntemlerinin (Ortalama Görüş Puanı (MOS) yordayıcıları v...
arXiv
GENCO – Kararlı Durum Izgara Analizi için Geliştirme Çerçevesine Gömülü Birleşik Bir Sinir Çözücü
Temel modeller iş iş akışlarını dönüştürüyor ve üretkenliği artırıyor, ancak katı fiziksel tutarlılı...