8B

In the model registry

Timeline

1. Oct 18, 2026 · Model release · 6 sources
   deepseek-ai/DeepSeek-V4.1-Flash-1017
   deepseek-ai published the model DeepSeek-V4.1-Flash-1017 on Hugging Face.
2. Oct 17, 2026 · Tutorial / explainer · 14 sources
   Local and Distributed Large LLM Inference with MLX and RDMA on Mac Studio
   In this post, we show how to run the latest frontier-class model — Muse Spark Open, with nearly a trillion parameters — right from your desk, using the latest version of MLX, released today.
3. Oct 17, 2026 · Opinion / analysis · 1 source
   [AINews] High Taste Testers Ep. 1
   We are piloting a new show on Latent Space: High Taste Testers, a variant of our food show where we pull you in to a typical lunch conversation in San Francisco.
4. Oct 17, 2026 · Opinion / analysis · 1 source
   llama.cpp has come a long way
   I started out back in March 2025 running Mistral Nemo Q4KM with 8K F16 on an AMD Radeon RX 7800 XT 16GB, with first ollama and then lmstudio.
5. Oct 16, 2026 · Research paper · 1 source
   FuTransformer: FP8 prefill and INT4 decode for Llama-style inference on Hopper
   We present FUTRANSFORMER, a CUDA implementation of a Llama-style Transformer block for NVIDIA Hopper GPUs.
6. Oct 16, 2026 · Research paper · 1 source
   Recognizing Study Lineage Is Not Enough: Over-Crediting Same-Trial Evidence in LLM-Based Medical Verification
   A model can identify shared trial lineage correctly in a direct task yet over-credit a second publication from the same trial when aggregating evidence.
7. Oct 16, 2026 · Open-source release · 1 source
   axolotl-ai-cloud/axolotl v0.21.0
   This release is headlined by Reflection's Beam, a 501B-parameter MoE you can QLoRA-tune on a single node.
8. Oct 16, 2026 · Research paper · 1 source
   Inferring L1 Backgrounds from L2 English with LLM Token Probabilities
   Native language identification (NLI) infers a non-native speaker's first language (L1) from their second language (L2) production.
9. Oct 16, 2026 · Research paper · 1 source
   Agentic Image Generation in the Wild
   Agentic image generation pairs an image generator with a search agent that retrieves references from the web, extending generation beyond the generator's training data.
10. Oct 16, 2026 · Research paper · 1 source
    Chinese Decision: A Family of Language Models for Single-Pass Probabilistic Decision-Making
    We present Chinese Decision, a family of Chinese-language models for single-pass probabilistic decision-making.
11. Oct 16, 2026 · Research paper · 2 sources
    The Design Space of Tri-Modal Masked Diffusion Models
    Diverging from previous approaches, we introduce the first tri-modal Masked Diffusion Models (MDM) pretrained from scratch on text, image-text, and audio-text data.
12. Oct 16, 2026 · Research paper · 2 sources
    Towards Looped Models Done Right, Part I: Topology, Input Injection, Recurrent-State Design
    Looped language models reuse physical Transformer blocks across logical depth, but established architectures entangle recurrence placement, input injection, and recurrent state organization.
13. Oct 16, 2026 · Research paper · 1 source
    Many Biases, One Circuit: Post-Training Tunes a Pretrained Cue-Routing Mechanism
    Language models often abandon a correct answer when the prompt contains a cue, a hint toward a wrong option such as "I think the answer is B.
14. Oct 16, 2026 · Research paper · 1 source
    Batched Contextual Reinforcement II: Does RLVR Need Long Rollout Windows?
    Long rollout windows make reinforcement learning with verifiable rewards (RLVR) expensive, yet simply shortening them can remove the correctness signal: when every single-problem rollout truncates before an answer, group-relative accuracy advantages vanish.
15. Oct 16, 2026 · Opinion / analysis · 1 source
    ALoDLM 8b looped diffusion language model: ~3x inference speed vs its base model qwen3-8b with better benchmark results
    Diffusion language models are the future!
16. Oct 16, 2026 · Research paper · 1 source
    VTrajEmbed: Aligning Text with Video Object Trajectories via Spatiotemporal Embeddings
    To address this issue, we propose VTrajEmbed, an embedding model that decomposes a video into object trajectories and aligns each trajectory embedding with text for object-level grounding through retrieval.
17. Oct 16, 2026 · Research paper · 1 source
    Reasoning Reranking with Semantic IDs as the Sole Item Representation
    Reasoning rerankers write a rationale before ranking a candidate slate, and existing systems read item metadata to do so, even though generative recommenders increasingly represent items as compact Semantic IDs (SIDs).
18. Oct 15, 2026 · Opinion / analysis · 1 source
    nvidia/Nemotron-3.5-Embed-1B-BF16 · Hugging Face
    Nemotron-3.5-Embed-1B-BF16 is a versatile embedding model developed by NVIDIA for multimodal and multilingual question-answering retrieval.
19. Oct 15, 2026 · Research paper · 1 source
    EgoP$^{2}$C: Internalizing Plan-Perceive-Compose for Spatial Reasoning in Egocentric Video
    Answering complex spatial reasoning questions from egocentric video requires determining what evidence is needed, acquiring targeted observations, and combining them through multiple reasoning steps.
20. Oct 15, 2026 · Opinion / analysis · 1 source
    Ternary Bonsai 2 Family: 2B, 4B, & 9B based on Qwen 3.5
    Going to try these out to see if they have the same looping behaviors that other Bonsai models have had.
21. Oct 15, 2026 · Model release · 2 sources
    Granite Switch 4.1: Chat, Decisions, and Speech from One Endpoint
    Specifically, Granite-Switch-4.1-8B and -30B models are configured to include:
22. Oct 15, 2026 · Research paper · 1 source
    DK-VLM: Geometry as Position for Spatial Reasoning in Vision-Language Models
    Vision-language models still struggle to understand 3D spatial relations across views and update spatial judgments as the observer moves.
23. Oct 15, 2026 · Research paper · 1 source
    KnowGPO: Knowledge-Grouped Policy Optimization for Search Agents
    We propose KnowGPO, a reinforcement learning framework that aligns trajectories by the task-relevant knowledge they have acquired rather than by their surface observations.
24. Oct 15, 2026 · Research paper · 1 source
    Training Reasoning Models to Generate Diverse Scientific Ideas
    To optimize for this, we propose training with a set-level reinforcement learning objective.
25. Oct 15, 2026 · Research paper · 2 sources
    Intern-Decision: Efficient Multimodal Decision Making with Autoregressive Masked Language Modeling
    We present Intern-Decision, a family of 0.8B, 2B, and 4B multimodal decision models that return structured choices and candidate probabilities from text and optional images.
26. Oct 15, 2026 · Research paper · 1 source
    TIPS: A Tiny Inference-time Policy for Scaling in Diffusion Language Models
    Masked diffusion language models expose an inference-time choice: which positions become visible after each denoising pass.
27. Oct 15, 2026 · Research paper · 1 source
    Effectiveness of Hidden-State Probes for Localizing Reasoning Failures
    We test the hidden states of 4B and 8B reasoning models for the failure type and the failing step and compare each probe with the strongest baseline that uses no hidden states.
28. Oct 15, 2026 · Research paper · 1 source
    Anvil: Scaling Embodied Spatial Intelligence through Knowledge Recirculation
    We introduce Anvil, a unified framework for data generation, training, and benchmarking for embodied spatial intelligence.
29. Oct 15, 2026 · Research paper · 1 source
    RoleQ: Role-Aligned Scales for Converting Language Models to Ternary
    We introduce RoleQ, a quantization framework that places scales by the functional roles of Transformer units rather than by the layout of the weight matrix, as conventional formats do.
30. Oct 15, 2026 · Model release · 12 sources
    Gemini 4 Argon: Built to solve your hardest problems
    Gemini 4 Argon: Built to solve your hardest problems

Often appears with