最新模型

新模型、開放權重、多模態、語音、影像與模型架構。

SPP Instills Value Reflection From the First Token, Widening the 3B Model’s OOD Risk Gap to 19 Points

Synthetic Persona Pretraining embeds first-person reflections generated from a “values constitution” into roughly 10% of pretraining documents, then uses post-training to bind that persona to the assistant identity. Newly released full experiments, weights, and data show that early intervention mainly improves value generalization, while resisting conventional jailbreaks may not require training from scratch.

2026-08-15

推論系統

RMM Prunes the Matrix Multiplication Contraction Axis Based on Input, Accelerating Llama 3.1 8B Long-Context Inference by Up to 1.40×

Reduced Matrix Multiplication leaves weights unchanged and, based on the column norms of the current activation, computes only the highest-scoring shared dimensions in a matrix multiplication. A100 tests show that gains increase with sequence length, but for short contexts, selection and kernel-launch overhead almost entirely offset the savings.

2026-08-15

最新模型

PlayWorld Has Multimodal Agents Control Nine World Models, but the Highest Long-Horizon Consistency Score Is Still Just 2.12/5

PlayWorld uses an agent that observes generated frames and dynamically adjusts control commands, preventing differences in movement magnitude across world models from making fixed trajectories incomparable. Results from 171 scenarios show that although current models can generate fluid visuals, they still struggle to preserve global spatial structure and states that continue evolving outside the field of view.

2026-08-15

AI 基礎設施

OpScale Breaks LLM Scaling Down to the Operator Level, Cutting Average GPU Use for Qwen2 Serving by 35%–50%

Instead of replicating an entire model, OpScale scales only the operators currently creating bottlenecks—such as attention and MLP—based on traffic and sequence length. The authors replayed 929,000 requests on A100 and GB200 clusters, using up to 36.3% fewer GPUs while meeting the same latency targets, but the prototype code has not yet been released.

2026-08-15

最新模型

TsuGO Reconstructs Search Trees from Reasoning Text; Gemini 3.1 Pro Scores Just 19% on Open-Ended Hard Problems

Using 600 verifiable Go life-and-death problems, TsuGO measures whether models allocate reasoning resources to the correct branches rather than merely counting correct answers and tokens. After four multiple-choice candidates are removed, both accuracy and search efficiency fall sharply across several reasoning models, leaving a clear gap with the neural-network-guided KataGo.

2026-08-15

最新模型

Post-Norm Beats Pre-Norm During Progressive Transformer Depth Growth, Cutting CE by 0.0328 in a 9-Layer Student Model

A Qwen3-8B distillation experiment found almost no difference between the two normalization placements when training the full depth jointly, but Post-Norm clearly outperformed after Transformer blocks were added in stages. The results show that normalization placement should not be chosen independently of the model growth strategy.

2026-08-15

最新模型

LongEarth-R1 Reasons Over Satellite Sequences of Up to 30 Frames, Outperforming Comparison Models Across All 12 Spatiotemporal Tasks

LongEarth-Bench organizes 117,000 remote-sensing images into approximately 120,000 long-sequence question-answer pairs, requiring models to track changes in landforms, anomalies, and missing states. LongEarth-R1, built on Qwen2.5-VL-7B, adds frame numbering, structured reasoning, and spatiotemporal rewards, but its data, weights, and code have not yet been released.

2026-08-15

最新模型

V-RAE Uses Frozen Visual Representations as Video Latents, Converging Up to 6× Faster in K600 Generation Training

Instead of training a video VAE solely for pixel reconstruction, V-RAE adds temporal compression and a decoder on top of semantic features from models such as DINOv3, SigLIP2, and V-JEPA 2.1. The authors found that the new latents improve both generation and future-frame prediction, though they have so far demonstrated only low-resolution, short-clip, class-conditional generation.

2026-08-15

最新模型

Mimir 1B Replaces Deep Stacking With Hierarchical Recurrence, Outperforms Tested 9B Models on Average Across Danish Benchmarks

The Danish Foundation Models team trained Mimir v1 from scratch, using HRM’s high- and low-level recurrent loops to approach some 4B models across the authors’ 20-benchmark evaluation. The weights are public, but “permissible data” does not mean fully open data: some corpora were used only under agreements or EU research exemptions.

2026-08-14

最新模型

“Test awareness” in the same set of models flips from positively to negatively correlated as prompts change, while the models themselves explain only 0.8% of the variance

A replication found that changing the probe’s contextual framing could shift the correlation with model scale from +0.94 to −0.82. The study recommends measuring across at least nine to ten prompt wrappers; otherwise, a single activation direction is insufficient for comparing models.

2026-08-14

推論系統

DARTree Builds Trees in Batches Before Pruning, Accelerating Qwen3-4B Decoding to as Much as 9.73× the Speed of Local AR

DARTree extends causal correction for diffusion draft models from a single path to multiple branches and replaces node-by-node heap search with batched expansion by depth. In tests on a single RTX 6000 Ada, it delivered higher average speedups across seven tasks than DFlash, DDTree, and Domino, but the code has not yet been released.

2026-08-14

最新模型

Intern-S2-Preview-397B Adds a Numerical Forecasting Branch, While a 4B Memory Module Raises the Average Biology Score by 3.40 Points

A new technical report brings scientific PDFs, time series, tool use, and long-horizon agent tasks into a unified training pipeline, while releasing the weights under Apache 2.0. The standalone Memory Decoder can add domain capabilities while keeping the 397B backbone frozen, but so far it has been demonstrated only with a biology module.

2026-08-14