lab DeepSeek (Hugging Face) · 503d ago

DeepSeek-V3-0324 improves reasoning, coding, and Chinese writing over DeepSeek-V3

from English

DeepSeek has released DeepSeek-V3-0324, a model update that demonstrates notable improvements over its predecessor, DeepSeek-V3, particularly in reasoning capabilities, code executability, and Chinese writing proficiency.

MMLU-Pro score increased from 75.9 to 81.2, GPQA from 59.1 to 68.4, AIME from 39.6 to 59.4, and LiveCodeBench from 39.2 to 49.2.
Enhanced front-end web development output with improved code executability and aesthetics.
Optimized Chinese writing style to align with R1 and improved medium-to-long-form content quality.
Fixed function calling accuracy issues present in previous V3 versions and enhanced multi-turn interactive rewriting.

The update provides better benchmark performance and more reliable function calling for developers using libraries like Transformers, vLLM, or SGLang.

Importance 3/3 DeepSeek (Hugging Face) DeepSeek Open weights Reasoning models

Benchmarks

Benchmark	Model	Score
MMLU-Pro	DeepSeek-V3-0324	81.2%
GPQA Diamond	DeepSeek-V3-0324	68.4%
AIME 2024	DeepSeek-V3-0324	59.4%
LiveCodeBench	DeepSeek-V3-0324	49.2%

Read original

arxiv arXiv cs.AI · 41d ago

RaDaR: A specialized reasoning LLM for accelerating rare disease diagnosis

Researchers present RaDaR, an open-source 32B parameter reasoning large language model designed to accelerate the diagnosis of rare diseases by addressing challenges in clinical deployability and data scarcity. The model was trained on nearly 50,000 public cases and over 100,000 synthetic cases, demonstrating superior performance across benchmarks and external validation centers.

arxiv arXiv cs.LG · 53d ago

Baseline Evaluation of Open-Source LLMs for Multi-Label ATT&CK Classification

A ground-truth dataset of 2,076 human-annotated sentences from 83 complex CTI reports was constructed and mapped to 114 ATT&CK techniques with \k{appa} = 0.68 inter-annotator agreement. Seven open-source LLMs ranging from 8B to 236B parameters were evaluated, achieving a maximum micro-averaged F1 score of 0.22. Parameter size showed a statistically significant positive correlation with F1 score, while prompt strategy and temperature did not yield significant improvements, indicating current open-source LLMs are insufficient for production-grade ATT&CK classification.

arxiv arXiv cs.CL · 53d ago · 2 views

LLMs Predict Dementia and Depression from Clinical Speech

A study uses open-weight large language models to assess dementia and depression severity from clinical interviews. LLMs achieve accurate zero-shot depression prediction (MAE 0.60) and improved dementia assessment with feature extraction (MAE 0.78), reducing errors by up to 35%. Pause-enriched transcripts match human transcriptions, supporting automated screening pipelines for neuropsychiatric disorders.

lab DeepSeek (Hugging Face) · 438d ago

DeepSeek-R1-0528 improves reasoning and AIME accuracy to 87.5%

DeepSeek has released DeepSeek-R1-0528, a minor version upgrade that significantly enhances depth of reasoning and inference capabilities through increased computational resources and algorithmic optimizations during post-training.

media InfoQ · 551d ago

DeepSeek open-sources DeepSeek-R1 LLM with performance comparable to OpenAI's o1

DeepSeek has open-sourced DeepSeek-R1, a large language model fine-tuned using reinforcement learning to enhance reasoning capabilities. The model achieves results on par with OpenAI's o1 on benchmarks like MATH-500 and SWE-bench.