Papersby GlyphDeepSeek-V3.2: Pushing the Frontier of Open Large Language Models research paper by DeepSeek, 2025
DeepSeek · Agents and evaluation · Dec 2, 2025 · 887 citations · 275 upvotes · unverified
What it shows
DeepSeek-V3.2 introduces DeepSeek Sparse Attention and a scalable reinforcement learning framework, achieving superior reasoning and performance compared to GPT-5 and Gemini-3.0-Pro in complex reasoning tasks.
By DeepSeek-AI, Aixin Liu, Aoxue Mei and 261 more · arXiv 2512.02556 · PDF
UnverifiedHugging Face's summary; not yet checked by hand.
More from DeepSeek
All 22DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache CompressionA 552B mixture-of-experts model with a 1M-token context, built to shrink the KV cache for agent workloads.
DeepSeek158 citations188 upvotesRead →DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive GenerationDSpark enhances LLM inference speed by combining parallel draft generation with adaptive verification that reduces waste and improves throughput in high-concurrency settings.DeepSeek77 citations49 upvotesRead →DualPath: Breaking the Storage Bandwidth Bottleneck in Agentic LLM InferenceDualPath addresses KV-cache storage I/O bottlenecks in multi-turn LLM inference by introducing dual-path loading and dynamic load balancing across prefill and decode engines.DeepSeek22 citations55 upvotesRead →DeepSeek-OCR 2: Visual Causal FlowDeepSeek-OCR 2 introduces DeepEncoder V2 that dynamically reorders visual tokens based on semantic content, enabling more human-like causal reasoning in 2D image understanding through cascaded 1D causal structures.DeepSeek79 citations73 upvotesRead →mHC: Manifold-Constrained Hyper-ConnectionsManifold-Constrained Hyper-Connections (mHC) stabilize and scale residual connection architectures by restoring identity mapping properties through manifold projection and infrastructure optimization.DeepSeek108 citations334 upvotesRead →DeepSeekMath-V2: Towards Self-Verifiable Mathematical ReasoningA self-verifying large language model for theorem proving improves mathematical reasoning by incentivizing rigorous step-by-step derivations and achieves high scores in international competitions.DeepSeek83 citations98 upvotesRead →DeepSeek-OCR: Contexts Optical CompressionDeepSeek-OCR uses optical 2D mapping to compress long contexts, achieving high OCR precision with reduced vision tokens and demonstrating practical value in document processing.DeepSeek200 citations98 upvotesRead →Insights into DeepSeek-V3: Scaling Challenges and Reflections on Hardware for AI ArchitecturesDeepSeek-V3 addresses hardware limitations through MLA, MoE, FP8 training, and Multi-Plane Network Topology, enabling efficient large-scale LLM training and inference.DeepSeek116 citations77 upvotesRead →
Other agents and evaluation papers
TopicGemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic CapabilitiesGemini 2.X model family, including Gemini 2.5 Pro and Flash, offers superior coding, reasoning, and multimodal understanding capabilities across a range of computational efficiencies.
Google4,594 citations67 upvotesRead →WorkArena: How Capable Are Web Agents at Solving Common Knowledge Work Tasks?WorkArena and BrowserGym evaluate large language model-based agents' ability to perform enterprise software tasks, revealing gaps in current agent capabilities and differences between open and closed-source LLMs.
ServiceNow443 citations2 upvotesRead →Aya Model: An Instruction Finetuned Open-Access Multilingual Language ModelAya, a multilingual generative language model supporting over 50% lower-resourced languages, excels in both generative and discriminative tasks across 99 languages, and provides extensive evaluations and open-source resources.
Cohere438 citations48 upvotesRead →Rubrics as Rewards: Reinforcement Learning Beyond Verifiable DomainsRubrics as Rewards (RaR) framework uses structured rubrics as interpretable reward signals for on-policy training, improving performance in real-world reinforcement learning tasks with subjective criteria.
Scale AI431 citations6 upvotesRead →SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?SWE-Bench Pro is a challenging benchmark for coding models, featuring complex, enterprise-level problems that require substantial code modifications, with performance evaluations showing significant limitations in current models.Scale AI367 citations21 upvotesRead →A Careful Examination of Large Language Model Performance on Grade School ArithmeticEvaluation of large language models on a new math benchmark, GSM1k, reveals that many models exhibit signs of overfitting to the existing GSM8k benchmark, leading to performance drops on GSM1k.Scale AI255 citations33 upvotesRead →
About this paper
- Venue
- arXiv.org
- Citations
- 887, 83 influential · Semantic Scholar
- Upvotes
- 275 · Hugging Face
- Lab
- DeepSeek · on Companies · on breaches · on Rounds · on labs · on Leaders · on Arena · on legislations · on Paydays · on Metered · on problems · on repos · on socials · on standards · on terms
Changes
Hugging Face upvotes rose from 273 to 275.
Influential citations rose from 82 to 83.
Citations rose from 805 to 887.
Influential citations rose from 80 to 82.
Citations rose from 784 to 805.
Influential citations rose from 77 to 80.
Citations rose from 766 to 784.
Influential citations rose from 74 to 77.
Citations rose from 761 to 766.
A first count of 74 influential citations.
A first count of 761 citations.
Found by the weekly Hugging Face scan; not yet checked by hand.