Sumi is a 7B-parameter uniform diffusion language model pretrained from scratch on 1.5T tokens. It competes with autoregressive models on knowledge, reasoning, and coding tasks but underperforms on commonsense benchmarks, likely due to its education-heavy data mixture. The model weights, checkpoints, and full training recipe are publicly released.
arxiv
arXiv cs.CL
·
84d ago
·
research
Sumi: Open Uniform Diffusion Language Model from Scratch
from English
Related articles
blog
Simon Willison
·
58d ago
·
3 views
Simon Willison creates custom Codex Desktop pet using GPT-5.6 and gpt-image-2
Simon Willison demonstrates how to create a custom animated "pet" for the Codex Desktop application by leveraging OpenAI's GPT-5.6 Sol model and the gpt-image-2 API.
arxiv
arXiv cs.AI
·
83d ago
·
6 views
AutoPass: Evidence-Guided LLM Agents for Compiler Performance Tuning
AutoPass uses runtime and compiler evidence to guide LLM-generated optimization decisions, outperforming expert heuristics and classical autotuning methods. It achieves geometric-mean speedups of 1.043x on x86-64 and 1.117x on ARM64 systems without prior training or fine-tuning.
arxiv
arXiv cs.LG
·
83d ago
·
4 views
LLM-Generated GPU Kernels Face Correctness Illusion
Benchmarks using fixed-shape checks miss real bugs in LLM-generated GPU kernels. A controlled corpus of 24 kernels, including 9 buggy variants with transcription errors, reveals that an op-schema-aware oracle detects all failures and passes all correct controls, with identical results across five GPU architectures.
arxiv
arXiv cs.CL
·
83d ago
·
4 views
AgentFinVQA: Auditable, On-Premise Financial Chart QA
AgentFinVQA introduces a multi-agent pipeline for financial chart question answering that ensures auditability and on-premise deployability without significant accuracy loss. It outperforms baseline models by +7.68 pp using a proprietary backbone and +4.84 pp with open-weights Qwen3.6-27B-FP8, while providing a confidence signal via verifier output that improves human review routing.
arxiv
arXiv cs.CL
·
83d ago
·
2 views
Zero-Shot Agentic LLMs Extract Lung Pathology from Narratives
A zero-shot agentic workflow using open-source LLMs extracts 13 College of American Pathologists synoptic fields from lung resection pathology reports. The best model (GPT-OSS-20B) achieved a Micro-F1 of 0.893, outperforming baseline recall and accurately capturing complex pathologic relations without task-specific training.