ZeroHour
DashboardNewsCVEsExploitsIncidentsAIBriefAnalyticsEntitiesWatchlistIOCsAsk the feedSourcesFeeds & APIKiosk mode

Search: “mixture-of-experts”

11 stories in the last 7d

Searchkeywordsmart Sortlatesthotimportancerelevance
DomainAllCyberAIWindow24h3d7d30dAllIn the wildImportant onlyShow duplicates
CategoryAllAI research 7AI industry 2Model release 2AI tools & infra 1
SeverityAllcritical 0high 0medium 0low 0info 12

Higher-order pruning of experts in mixture-of-experts language models

New HOPE method uses second-order objectives to prune Mixture-of-Experts LLMs, outperforming REAP at 50% pruning on models up to 122B parameters.

Researchers derive HOPE (Higher-Order Pruning of Experts), a second-order pruning objective for Mixture-of-Experts language models that provably minimizes an upper bound on pruning error by modeling cooperative expert interactions. They show REAP, a state-of-the-art first-order method, is a special case of HOPE with interaction terms ignored. Across three frontier MoE models up to 122B parameters, two calibration sets, and benchmarks covering math, instruction following, coding, and agentic tasks, HOPE achieves the best average rank (1.58 of 5 at 50% pruning versus 2.42 for REAP), with gains up to +6.1% on agentic coding.

arXiv cs.AI / cs.LG / cs.CL · 20h agoAI research

Flattening Every Memory Peak in Long-Context Mixture-of-Experts Training

Four scheduling techniques bound all memory peaks in long-context MoE training, enabling 120B-667B models at 1M-token context with up to 10.4x throughput.

The paper addresses memory peaks in long-context Mixture-of-Experts training by bounding four unbounded components: expert dispatch with the routing matrix, vocabulary projection, gradient checkpoint boundaries, and optimizer state. It introduces PipelinedLLEP (capped token contributions per dispatch chunk), Ring-DTP (ring circulation of activations or weight shards with online log-sum-exp), Selective Checkpoint Offload (SCO), and OffloadStreamAdamW, all preserving exact loss and gradients. Composed on MoE models from 120B to 667B parameters, the methods enable training at 1M context length, 8-32x the context reach of a tuned FSDP2 baseline, and up to 10.4x its throughput.

Hugging Face daily papers · 4d agoAI research1

Data Scarcity and Model Sparsity: Mixtures-of-Experts Overfit More to Repeated Data

Study finds Mixture-of-Experts models overfit faster than dense Transformers under repeated training data, with degradation tied to total parameter sparsity.

Across models from 80M to 1B active parameters (8.5B total), MoE architectures degrade more rapidly than dense models when training data is repeated, with the effect increasing with sparsity as dictated by total parameters. Dense 80M models tolerate 8x repetition with minimal loss while MoEs suffer at 4x and underperform dense models beyond 32x. Masking-based regularization such as dropout mitigates overfitting, letting MoEs beat dense models even at over 64x repetition, though no method matches all-unique training data. Routing stabilizes early and expert specialization correlates with overfitting to repeated data.

arXiv cs.AI / cs.LG / cs.CL · 6d agoAI research

DeepSeek-v4.1 Flash: Pushing the Limits of KV Cache Compression

DeepSeek-V4.1 Flash is a 552B-parameter multimodal MoE model with 1M-token context achieving 4x KV cache compression for long-horizon agent workloads.

A detailed analysis of the DeepSeek-V4.1 Flash technical report describes a 552B-parameter multimodal mixture-of-experts model supporting contexts up to 1 million tokens. Its Causal Encoder-Decoder (CED) architecture activates 8B parameters during prefill and 16B during decode, and reportedly delivers about 420 tokens/s. Joint optimization of architecture (CSA2 cross-layer compression), FP4 KV cache precision, and deployment strategy cuts runtime KV cache to roughly 1/4 and persistent KV cache to about 1/8 of DeepSeek-V4-Flash at the same sequence length, targeting storage and bandwidth bottlenecks in long-horizon agent serving. The author notes all DeepSeek-V4 Pro models were taken offline following the release.

Hacker News · AI · 11h agoModel releaseHN 41↑ · 3 comments

NVIDIA Vera Rubin NVL72 Delivers Leading Performance in MLPerf Inference v6.1 Debut

NVIDIA's Vera Rubin NVL72 debuts in MLPerf Inference v6.1 with up to 3.7x higher throughput than GB300 NVL72 and 99% scaling efficiency at 288 GPUs.

In its first MLPerf Inference preview submission, NVIDIA's Vera Rubin NVL72 achieved up to 3.7x higher throughput than GB300 NVL72 on Qwen3-VL and 2.5x on DeepSeek-R1. A 288-GPU GB300 NVL72 submission across four racks reached 99% scaling efficiency on the DeepSeek-R1 offline benchmark. Software optimizations delivered up to 1.6x gains over v6.0, leveraging TensorRT-LLM, vLLM, Dynamo, disaggregated serving, and NVFP4 precision.

NVIDIA Blog · 22h agoAI industry

The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Predictionnew

Edge0 streams 35B-parameter MoE inference from SSD on a 24GB machine at 20 tokens/s using a one-token-ahead prerouter, with framework and checkpoints open-sourced.

The paper (arXiv 2609.18063) presents Edge0, a streaming MoE inference engine whose per-layer prerouter predicts the next layer's expert routing one token ahead so SSD reads can be pre-staged with nothing dropped. An unmerged recovery LoRA trained on the student path compensates for quality lost to int4 quantization and routing replacement. On a single 24GB machine it serves a 35B MoE at 20 tok/s within 3GiB of peak active memory, within a few points of its fp16 teacher across five public benchmarks; framework, checkpoints, and adapters are open source.

Hugging Face daily papers · 1d agoAI tools & infra

What Breaks Under Pruning in Smart Homes, and When? Evaluating LLM Degradation Across Architectures and Task Complexity

Pruning study across four LLM architectures finds dense models degrade sharply on smart-home tool calling while MoE models tolerate far more.

Researchers systematically study pruning-induced degradation in smart-home tool calling across four LLMs spanning dense Transformer, dense hybrid, and mixture-of-experts architectures, combining depth, width, hybrid, and expert pruning methods, and evaluate over 19,500 instances from three datasets after post-pruning supervised fine-tuning. Dense models show narrow safe pruning regions followed by sharp degradation, while MoE models tolerate substantially more pruning. Pruning degrades grounded specificity (operation, device, argument, value) before schema-level intent, and aggressive dense pruning can induce systematic over-refusal.

arXiv cs.AI / cs.LG / cs.CL · 1d agoAI research

Perplexity Portable Computer Is Now Available on Windows, Powered by NVIDIA RTX

Perplexity launches Portable Computer local AI agent on Windows for NVIDIA RTX PCs with 24GB+ VRAM, keeping sensitive work on-device.

Perplexity released Portable Computer, a local version of its agentic Perplexity Computer, in its Windows app for NVIDIA GeForce RTX PCs and RTX PRO Workstations with 24GB or more VRAM. It runs a locally post-trained model such as Qwen 3.8 27B optimized for NVIDIA RTX GPUs, handling multistep tasks and file analysis on-device with a SPACE sandbox and built-in browser. Connectors cover Outlook, OneDrive, Word, Google Drive, Gmail, Slack, and GitHub, and the agent can escalate to cloud models only with user permission.

NVIDIA Blog · 2d agoAI industry

Cohere Releases North Small Translate: A 218B MoE Translation Model That Scores 83.6 on WMT26 Across 50 Languages

Cohere released North Small Translate, an open-weight 218B MoE (25B active) translation model scoring 83.6 on WMT26 across 50 languages.

Cohere and Cohere Labs released North Small Translate, a decoder-only sparse Mixture-of-Experts translation model with 218B total and 25B active parameters, 128 experts with 8 activated per token plus shared experts, and 16K-token input and output context. In Cohere's vendor-reported WMT26 evaluation, judged by GPT-5.6-Sol, it scores 83.6 averaged across 50 languages (84.36 in an agentic multi-pass mode), ahead of DeepL NextGen (81.37), Qwen 3.5 397B A17B (81.56), GLM 5.2 (76.50), and Google Translate (68.20). The model was built with RWS's Language Weaver team, post-trained specifically for translation, and reports 112 output tokens per second versus 81 for Gemma 4 31B, with long-document xCOMET-XL scores of 48.9 versus 21.3 for Google Translate. It is available free on Cohere's Chat V2 API until rate limits, with three self-hosting checkpoints including a 4-bit NVFP4 variant running on 1x B200 or 2x H100.

MarkTechPost · 6d agoModel release

Expert-Space Exploration in MoE Reinforcement Learning

ESRL explores MoE expert-routing space during RL post-training, improving Qwen3-30B-A3B Pass@1 by 3.2 points over GRPO without extra compute.

The paper shows perturbing expert routing increases rollout diversity similarly to higher decoding temperature, but naive perturbation degrades quality. ESRL anchors high-confidence experts, restricts stochastic routing to a plausible candidate pool, adapts perturbation strength via router entropy, and replays recorded expert paths during policy optimization. It achieves the best results across top-K, top-1, and shared-expert MoE backbones on math, science, and code tasks; on Qwen3-30B-A3B it improves average Pass@1 and Pass@8 over GRPO by 3.2 and 4.5 percentage points.

Hugging Face daily papersupdated · 5d agofirst · 6d agoAI research 2 sources

StepAudio 3 Music Technical Report

StepAudio 3 Music introduces long-form text-controlled music generation using ABC-notation planning and flow-matching diffusion, ranking near the top music arena.

StepAudio 3 Music generates long-form, text-controlled music using a 50-Hz single-codebook tokenizer with 65,536 entries and a flow-matching diffusion Transformer over VAE latents. A Mixture-of-Experts autoregressive model first plans an arrangement in ABC notation (ABC-CoT) before predicting music tokens. With DPO fine-tuning, it tops AudioBox content and production quality scores and reaches Quality Elo 1105 on the Artificial Analysis Music Arena, behind Suno V5.5 and Mureka. Generation covers songs, accompaniment from dry vocals, and cover synthesis up to 5 minutes 30 seconds at 48-kHz output.

Hugging Face daily papers · 6d agoAI research