AaronFeng47

• Kimi K2.6 is undergoing pilot testing

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 1 comments

• Gemma4-31B Matches Qwen 3.5-27B in Private Benchmark (with Much Lower Token Usage)

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 0 comments

• Jackrong/Qwopus3.5-9B-v3 · Hugging Face

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 0 comments

• I just realized Qwen3-30B-A3B is all I need for local LLM

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 231 comments

• New Qwen3.5 models spotted on qwen chat

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 198 comments

• Qwen 3.5 Plus(397b-a17b) is now available on Chinese Qwen APP

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 22 comments

• MNN Chat Android App by Alibaba

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 22 comments

• AM-Thinking-v1

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 17 comments

• Faster open webui title generation for Qwen3 models

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 14 comments

• Qwen3-30B-A3B GGUFs MMLU-PRO benchmark comparison - Q6_K / Q5_K_M / Q4_K_M / Q3_K_M

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 43 comments

• Ling-1T

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 91 comments

• ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 1 comments

• Critique Fine-Tuning: Learning to Critique is More Effective than Learning to Imitate

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 3 comments

• Unsloth GGUFs Perplexity Score Comparison | Qwen3-Coder-30B-A3B-Instruct

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 45 comments

• Huihui-gpt-oss-120b-BF16-abliterated

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 16 comments

• Gemma 3 27b now available on Google AI Studio

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 86 comments

• FlashMLA - Day 1 of OpenSourceWeek

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 94 comments

• OpenAI prompt optimizer

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 2 comments

• Interactive Reasoning Benchmarks | ARC-AGI-3 Preview

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 7 comments

• Baichuan-M2-32B / Medical-enhanced reasoning model

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 6 comments

• The OpenAI Open weight model might be 120B

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 172 comments

• Artificial Analysis Long Context Reasoning (AA-LCR) benchmark

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 1 comments

• 8% -> 33.3% on Aider polyglot

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 23 comments

• Serious hallucination issues of 30B-A3B Instruct 2507

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 24 comments

• MediPhi-Instruct

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 38 comments

• Private Eval result of Qwen3-235B-A22B-Instruct-2507

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 14 comments

• MetaStone-S1-32B

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 4 comments

• The 1T Kimi K2 model is using DeepSeek V3 architecture

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 30 comments

• What LLM Workflow UI Are You Using?

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 3 comments

• GLM-4.1V-Thinking

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 48 comments

• Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 5 comments

• Is this the largest "No synthetic data" open weight LLM? (142B)

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 48 comments

• FairyR1 32B / 14B

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 11 comments

• Unsloth's Qwen3 GGUFs are updated with a new improved calibration dataset

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 101 comments

• Cosmos-Reason1: Physical AI Common Sense and Embodied Reasoning Models

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 1 comments

• AceReason-Nemotron-14B: Advancing Math and Code Reasoning through Reinforcement Learning

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 3 comments

• A per-project memory feature for local models?

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 6 comments

• This merge is amazing: FuseO1-DeepSeekR1-QwQ-SkyT1-32B-Preview

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 70 comments

• Qwen: Parallel Scaling Law for Language Models

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 5 comments

• Qwen3-32B hallucinates more than QwQ-32B

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 37 comments

• Xiaomi MiMo - MiMo-7B-RL

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 18 comments

• OpenAI: Introducing HealthBench - An evaluation for AI systems and human health.

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 6 comments

• Qwen3-32B-Q4 GGUFs MMLU-PRO benchmark comparison - IQ4_XS / Q4_K_M / UD-Q4_K_XL / Q4_K_L

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 18 comments

• Absolute_Zero_Reasoner-Coder-14b / 7b / 3b

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 31 comments

• Make Qwen3 Think like Gemini 2.5 Pro

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 26 comments

• Absolute Zero: Reinforced Self-play Reasoning with Zero Data

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 8 comments

• Auto Thinking Mode Switch for Qwen3 / Open Webui Function

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 20 comments

• Qwen3-235B-A22B on livebench

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 35 comments

• Qwen3 on Dubesor Benchmark

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 10 comments

• Qwen3-32B-IQ4_XS GGUFs - MMLU-PRO benchmark comparison

  Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 48 comments