TurboMamba is a new language‑model architecture that combines a global Mamba‑3 model with a small Transformer to achieve faster inference, faster training, and better token prediction. It introduces three lightweight enhancements—direct access to the previous block’s last token, a small correction to next‑token scores, and a learned mixture with recent‑token frequencies—to improve local predictions. In experiments on FineWeb‑Edu, TurboMamba outperforms Transformer, Mamba‑3, Jamba, and Zamba, achieving the lowest mean test prediction loss and significantly higher training throughput and batch latency speedups.
By Hanxin Zhang
arXiv:2608. 20210v1 Announce Type: cross Abstract: Small language models are usually built like large ones and then squeezed onto a CPU afterwards.
By Christos Koutsiaris
The paper compares two low‑budget methods for converting a 30B Mixture‑of‑Experts autoregressive language model into a diffusion language model. One method updates a subset of the model’s weights in‑place, while the other freezes the context tower and conditions on a frozen causal copy via cross‑attention. With only 1B training tokens, the frozen‑tower approach achieves a HumanEval pass@10 score of 71.60 versus 6.19 for the in‑place method, and retains 95% of the parent’s GSM8K and 99% of its MMLU‑Pro performance.
By Wentao Lu, Jesse Clark, Tianyu Zhu
The paper introduces SafeStart, a technique for Mixture-of-Experts (MoE) language models that mitigates the safety degradation caused by active expert reduction during serving. By running all experts for the prompt and the first eight generated tokens before switching to a reduced set, SafeStart largely restores the model’s refusal to harmful requests while preserving most of the speedup achieved by expert reduction. Experiments on several MoE models show that SafeStart maintains high offline decode speedups (≈85% of the original) and limits the increase in harmful responses to within a few percentage points, even under jailbreak prompts.
By Erfan Baghaei Potraghloo, Saeid Shokoufa, Seyedarmin Azizi, Arya Fayyazi, Souvik Kundu, Mehdi Kamal, Massoud Pedram
Small language models are usually built like large ones and then squeezed onto a CPU afterwards. We did the opposite: we fixed the target first, one user, one token at a time, 4-bit weights, ordinary CPU, and chose the architecture to suit it.
arXiv:2608.30320v1 Announce Type: new
Abstract: We describe the architecture and ablations of Qwen3.8-Flash-Next, a sparse mixture-of-experts model with 125B parameters, 6B activated per token, and a...
By Zihan Qiu, Zekun Wang, Xiao Li, Yanpeng Li, Yang Xu, Yixuan Wang, Huaqing Zhang, Rui Men, Bochao Mao, Chengruidong Zhang, Fan Zhou, Hao Luo, Haofeng Huang, Haoran Lian, Haoyan Huang, Hongqing Chen, Jianwei Zhang, Jing Xu, Junjie Wang, Langshi Chen, Liangyu Wang, Linlang Jiang, Man Yuan, Minmin Sun, Peng Jin, Siqi Zhang, Siyu Wang, Xingzhang Ren, Yakai Wang, Yi Zhang, Yiming Dong, Yizhong Cao, Yubo Ma, Yunfei Mao, Bo Zheng, Dayiheng Liu
arXiv:2511. 20849v2 Announce Type: replace-cross Abstract: We introduce a new tokenizer for language models that minimizes the average tokens per character, thereby reducing the number of tokens needed to represent text during training and to generate text during inference.
By Dong Dong, Weijie Su
arXiv:2610.33025v1 Announce Type: cross
Abstract: Letting a language model take gradient steps on the text it is reading, known as dynamic evaluation, lowers its loss on the rest of the document. Its...
By Surya Saka
The paper reports that in English all‑words word sense disambiguation (WSD), the scarcity of high‑quality labels—not the models—has become the limiting factor. The authors introduce lexEN, a human‑adjudicated correction layer over the Maru2022 ALL_NEW benchmark, and SenseBench, a living leaderboard for LLM WSD evaluation. They show that frontier large language models reach about 95 % accuracy on lexEN‑v1, that relabeling corpora with these models improves downstream systems, and that fine‑grained WordNet senses are often ill‑posed, with coarsening improving both annotator agreement and model performance.
"whyItMatters":"The study highlights that improving label quality and managing annotation costs are now the critical challenges for advancing WSD performance, as model accuracy is already near its theoretical ceiling."
By Vassili Philippov, Amro Salman, Dmitrii Andreev, Penny Hands, Emil Kaiumov, Pavel Katunin, Anton Nikolaev
The paper demonstrates that a byte‑level BPE tokenizer can be sliced to create multiple vocabulary sizes from a single trained model, preserving exact logits while reducing deployed weights by 66%. Experiments on 30 models show that while sliced models match the full model numerically, they underperform fixed‑cap specialists by a few percentage points in bits‑per‑byte. Multi‑cap training improves robustness to typographical noise, suggesting benefits from training across multiple granularities rather than from control tokens alone.
By Christos Koutsiaris
arXiv:2606. 25156v3 Announce Type: replace-cross Abstract: Native length extrapolation remain a weakly solvable problem in language modeling due to trade-off balancing between exact retrieval fidelity, long-document likelihood, and inference efficiency.
By Habibullah Akbar
Kairos is a byte‑level foundation model that integrates bidirectional gated‑DeltaNet linear attention, sliding‑window attention, block‑wise Attention‑Residual aggregation, a sparse Mixture‑of‑Experts feed‑forward layer, and a multi‑scale byte‑level codec. It is trained with a uniform‑noise denoising objective over a single byte space shared by text, image, audio, video, lidar, and control streams, and the authors provide code‑derived algorithms for each component, including a cross‑session recurrent‑state memory gate. Convergence sanity checks on a 440‑byte text sentence show that the two attention variants reach a minimum loss near 10⁻³, and the implementation appears to optimize, though generalization and multimodal performance remain to be evaluated in future pretraining runs.
By Fabien Furfaro