AI Dataset2026
Provides the dataset and accompanying technical report for a DeepSeek project that interleaves spatial markers (points and boxes) into multimodal LLM reasoning. Includes a public subset of data and benchmarks under an MIT license; model weights are not included.
AI Dataset2026
Provides 19,331 multi-turn ChatML Hermes reasoning traces produced by DeepSeek V4 Pro for LoRA fine-tuning of agent-style models; includes VRAM-tiered variants, train/valid/test splits, and dense tool-calling annotations in Parquet format.
AI Dataset2026
Provides 19,331 multi-turn ChatML Hermes reasoning traces for LoRA fine-tuning of local models to behave as Hermes agents. Includes train/valid/test splits, VRAM-tiered variants (nano→spark), ~138K tool-call annotations, and Parquet format under Apache-2.0.
AI Dataset2026
Agentic coding evaluation dataset containing real-world, multi-step developer tasks and raw model responses across 20+ programming languages. Emphasizes challenging, persona-driven prompts for benchmarking and fine-tuning; users should filter and audit outputs before training.
Preview of an MoE model family (V4-Pro: 1.6T params, 49B active; V4-Flash: 284B, 13B active) built for 1M-token contexts. A hybrid attention design cuts single-token inference FLOPs to 27% and KV cache to 10% versus V3.2 at million-token length.
AI Dataset2026
Contains 4,006 newline-delimited JSONL agent-session traces recording assistant responses and tool calls from deepseek/deepseek-v4-pro — includes a training-ready tools schema snapshot and helpers for conversion to SFT/distillation workflows.
AI Model2026
Performs hour-scale video understanding and fine-grained temporal localization while exposing agent-style multimodal tool/code/search abilities. Built on a sparse-attention long-context architecture (DSA) and a specialized inference stack—best used in GPU-backed research or production evaluation.
Learns fine-grained preferences over sub-trajectories to identify and penalize redundant steps in long chain-of-thoughts, letting models "fold" reasoning chains into concise paths; reports ~56% token reduction on DeepSeek-R1-Distill-Qwen-7B while keeping accuracy.
AI Dataset2026
Pairs OCR-extracted annual-report text with ground-truth financial KPI values to benchmark LLM/table-QA and needle-in-a-haystack extraction tasks. Includes Markdown OCR (.mmd), page images for eval, and 31 KPI columns across multiple years—suited for KPI extraction, retrieval, and robustness testing.
AI Agent Papers2026
Synthesizes shortcut-resistant search tasks to train deep search agents by controlling four shortcut risks across entity selection, evidence-graph construction, question formulation, and adversarial refinement. Produces training trajectories with longer pre-answer search and fewer shortcut patterns; code will be released on GitHub.
AI Model2026
Provides a large language model optimized for long-horizon agentic tasks and end-to-end coding workflows — with a stable 1,000,000-token context, IndexShare sparse-attention and multi-level thinking-effort modes. MIT-licensed and designed for deployments that need sustained long-context reasoning and coding.
AI Model2026
Performs one-shot, long-horizon OCR and document parsing by using Reference Sliding Window Attention (R-SWA) to keep the decoder KV cache constant, enabling single-pass multi-page transcription; code, model weights and an accompanying arXiv report are provided.