model releases

[SYCL] support sparse FA (#28796) fix conflict fix format issue rm unused code Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/50188435 macOS/iOS: macOS App

rpc: include nb in the get_alloc_size cache key and floor the result at ggml_nbytes (#29283) rpc : include nb in the get_alloc_size cache key and floor the result at ggml_nbytes cont : remove redundan

HIP: bump HIP_VERSION requried for fp8 to avoid missing __hip_fp8_e4m3 support in 6.2 (#29231) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/50216799 macO

llama : add llama_prec_policy + model-driven W4A4 path (#24364) Rebase and update based on #26675 Signed-off-by: ynankani ynankani@nvidia.com CI failure fix(launh_bounds overload on HIP) and cleanup S

metal : split fa kernels into per-dtype libraries (#29329) metal : split fa kernels into per-dtype libraries Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp cont : minor fix comment Website: ht

metal: FWHT kernels for block widths above 512 (#29095) metal: FWHT kernels for block widths above 512 The Metal FWHT covers widths 64 to 512, one row per simdgroup with N/32 values per lane. Wider bl

common : extract shared unicode path/string helpers (#29415) Signed-off-by: Adrien Gallouët angt@huggingface.co Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestati

Fixing the vulkan build issue of legacy GLSLC version that has no cooperativeMatrix API support (#29373) (#29409) vulkan : fix build issue of legacy glslc version by adding GGML_VULKAN_COOPMAT_GLSLC_S

opencl: add bin kernel kernel_gemm_noshuffle_q5_k_f32_32b_trans_ila_a8_bin, kernel_gemm_noshuffle_q5_k_q8_1_dp4a_ila_a8_bin (#29401) opencl: add A8 Q5_K non-MoE non dp4a + dp4a binary kernel opencl: f

mtmd: fix mel preprocessor in LFM2 audio (#29403) which resulted in different greedy transcripts for 4.5% of English and 6.5% of Japanese test utterances. In Japanese, some differences changed entire

common,rpc : simplify fs_create_directory_with_parents() (#29432) The original function was broken on Windows for some unicode paths Paths without a trailing separator now create the last directory to

Banger paper from Microsoft and colleagues. We talk about human taste being important in this AI era. But for recursive self-improvement, an agent's taste also matters. The big question is: How often

arXiv:2609.29146v1 Announce Type: new Abstract: Bangla, the seventh most spoken language in the world, remains under-resourced for neural text-to-speech. Public Bangla speech corpora are dominated by

arXiv:2609.29371v1 Announce Type: new Abstract: This paper presents BanglaTurn, a corpus for end-of-turn detection in Bangla conversational speech, and a model trained on it. The corpus holds 35,374 s

arXiv:2609.29397v1 Announce Type: new Abstract: Ternary (1.58-bit) weights are attractive for microcontroller-class language models, but the sub-1M-parameter regime rests mainly on isolated, single-se

arXiv:2609.29266v1 Announce Type: new Abstract: We develop a multi-label Polish content-safety classifier by fine-tuning allegro/herbert-base-cased (124M) across five categories (hate, vulgarity, sexu

arXiv:2609.29798v1 Announce Type: cross Abstract: This paper presents an end-to-end study of automatic speech recognition (ASR) for adolescent health communication in three Ghanaian languages (Twi, Da

arXiv:2609.29559v1 Announce Type: new Abstract: Arabic-Russian machine translation (MT) remains under-explored due to the rich morphology of Arabic and low lexical overlap between the two languages. W

arXiv:2609.28673v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed as argumentative agents in persuasive dialogues, necessitating rigorous evaluation of their debat

Reinforcement learning (RL) has been a keystone of modern LLM post-training, but it demands large training clusters and access to model internals that external customers can't have with proprietary mo

arXiv:2609.29744v1 Announce Type: cross Abstract: We present: (i) a new dataset consisting of the full development history of a 21,000-line Python tool built entirely by Claude AI, with no human-autho

arXiv:2609.29960v1 Announce Type: cross Abstract: Fine-tuning large language models on new objectives can improve helpfulness, instruction following, or domain-specific performance, but it can also in

arXiv:2609.29866v1 Announce Type: cross Abstract: Deploying real-time speech enhancement on resource-constrained devices requires meeting strict latency, memory, and energy constraints. Microcontrolle

arXiv:2603.11133v2 Announce Type: replace Abstract: Sequence modeling tasks can involve intrinsic higher-order dependencies, while standard self-attention assigns scores to token pairs and does not ex

arXiv:2609.29934v1 Announce Type: new Abstract: Does progress on spatial reasoning benchmarks translate into better navigation? Existing benchmarks test isolated inferences from images or videos, with

arXiv:2609.28690v1 Announce Type: new Abstract: Faithful user simulation is fundamental to building, evaluating, and improving interactive AI at scale. However, plausible individual responses do not e

arXiv:2609.30223v1 Announce Type: new Abstract: Common segmentation losses aggregate errors voxel-wise, so lesions influence the objective in proportion to their volume, giving small but clinically cr

arXiv:2609.28758v1 Announce Type: cross Abstract: We describe our submission to the Unsupervised Speech in the Wild (UPS) Challenge at Interspeech 2026, a bidirectional Mamba-2 (BiMamba2) encoder trai

arXiv:2609.29069v1 Announce Type: new Abstract: We introduce BranchShine-CR, a 25M-parameter model for multilingual transcription into the International Phonetic Alphabet (IPA). It combines log-mel fe

Check out this week's updates and releases: — Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, two of our most expressive audio generation models yet — Gemini 3.8 Live with Live Avatar, bringing ne

arXiv:2609.29828v1 Announce Type: new Abstract: We present ChunkRank, an open-source Python library that derives chunk boundaries from a target model's tokenizer and context window, and selects an ans

arXiv:2609.28813v1 Announce Type: new Abstract: Cinematography, the craft of visual storytelling through framing, lighting, and camera operation, fundamentally shapes how audiences perceive and emotio

arXiv:2609.29479v1 Announce Type: new Abstract: Prospective clinical actions, the follow-ups, orders, referrals, and instructions that deter-mine what happens to a patient next, are annotated today in

arXiv:2609.30233v1 Announce Type: cross Abstract: Task and motion planning (TAMP) problems remain difficult even with full observability and object-centric states because discrete decisions are tightl

arXiv:2609.28826v1 Announce Type: new Abstract: Machine translation (MT) for Indian languages remains constrained by the limited availability of high-quality, Indic-centric parallel corpora and evalua

arXiv:2609.08646v2 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly deployed with hierarchical instructions, yet they remain vulnerable to conflicts in which user directi

arXiv:2609.29225v1 Announce Type: new Abstract: Lip synchronization aims to generate visual lip dynamics that align precisely with speech audio. Despite the high generation quality of diffusion models

arXiv:2609.29194v1 Announce Type: cross Abstract: Mobile robots require robust, real-time fault detection capable of continuous adaptation on constrained edge hardware. While deep time-series models e

arXiv:2609.28919v1 Announce Type: new Abstract: Harnesses, the products that run AI coding agents, are multiplying, and enterprises are rolling them out to their employees: what started as pilots with

A U.S. appeals court upheld the Trump administration’s decision to blacklist Anthropic’s Claude AI after the firm declined to enable certain features deemed critical for military operations. In a 2‑1

arXiv:2609.29075v1 Announce Type: new Abstract: Cancer registrars, including Oncology Data Specialists (ODSs), must interpret complex and frequently updated coding and staging standards. We developed

arXiv:2609.28972v1 Announce Type: cross Abstract: Cross-country recommendation on modern e-commerce platforms is typically deployed with disjoint user and item ID spaces across markets, removing the s

arXiv:2609.29711v1 Announce Type: cross Abstract: Privacy-preserving continual learning (PPCL) must reduce the reproduction of sensitive content while retaining useful knowledge across sequential task

arXiv:2609.28570v1 Announce Type: new Abstract: Reinforcement learning (RL) is widely used to sharpen reasoning in multimodal large language models (MLLMs), yet its effect on hallucination is uneven.

arXiv:2609.29509v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed as multi-turn agents that must sustain goals, use tools, and adapt to other agents over extended

arXiv:2609.29043v1 Announce Type: cross Abstract: General Purpose Service Robot (GPSR) tasks, as defined in the RoboCup@Home benchmark, require robots to interpret diverse natural language commands an