reasoning models

Alibaba’s new open source model QwQ-32B matches DeepSeek-R1 with way smaller compute requirements

• AI
• AI, ML and Deep Learning
• alibaba
• category-/Science/Computer Science
• China
• Conversational AI
• Deepseek R1
• GPUs
• inference
• Large Reasoning Models (LRMs)
• LLM reasoning
• NLP
• Nvidia
• o1
• o1-preview
• OpenAI
• openai o1
• Qwen
• Qwen 2.5
• qwen team
• qwen with questions
• qwq
• qwq-32b
• reasoning AI
• reasoning models
• Uncategorized

Alibaba’s new open source model QwQ-32B matches DeepSeek-R1 with way smaller compute requirements

admin March 6, 2025

While DeepSeek-R1 operates with 671 billion parameters, QwQ-32B achieves comparable performance with a much smaller footprint.Read More

Read More Read more about Alibaba’s new open source model QwQ-32B matches DeepSeek-R1 with way smaller compute requirements

OpenAI releases ‘largest, most knowledgable’ model GPT-4.5 with reduced hallucinations and high API price

• AI
• AI, ML and Deep Learning
• ChatGPT
• Conversational AI
• GPT 4o
• GPT-4
• gpt-4.5
• NLP
• o series
• OpenAI
• reasoning models
• Uncategorized

OpenAI releases ‘largest, most knowledgable’ model GPT-4.5 with reduced hallucinations and high API price

admin February 27, 2025

GPT-4.5 API pricing appears shockingly high, costing developers $75 and $180 for 1 million tokens in and...

Read More Read more about OpenAI releases ‘largest, most knowledgable’ model GPT-4.5 with reduced hallucinations and high API price

OpenAI releases GPT-4.5 claiming 10X efficiency over GPT-4, but says it’s ‘not a frontier model’

• AI
• AI, ML and Deep Learning
• ChatGPT
• Conversational AI
• GPT 4o
• GPT-4
• gpt-4.5
• NLP
• o series
• OpenAI
• reasoning models
• Uncategorized

OpenAI releases GPT-4.5 claiming 10X efficiency over GPT-4, but says it’s ‘not a frontier model’

admin February 27, 2025

The API pricing appears shockingly high, costing developers $75 and $180 for 1 million tokens in and...

Read More Read more about OpenAI releases GPT-4.5 claiming 10X efficiency over GPT-4, but says it’s ‘not a frontier model’

IBM Granite 3.2 uses conditional reasoning, time series forecasting and document vision to tackle challenging enterprise use cases

• AI
• category-/Computers & Electronics
• category-/Science
• Deepseek R1
• enteprise AI
• IBM
• IBM Granite
• open source AI
• reasoning models
• time series forecasting
• Uncategorized
• vision model

IBM Granite 3.2 uses conditional reasoning, time series forecasting and document vision to tackle challenging enterprise use cases

admin February 26, 2025

IBM is bringing the power of conditional reasoning to its open source Granite 3.2 LLM, in an...

Read More Read more about IBM Granite 3.2 uses conditional reasoning, time series forecasting and document vision to tackle challenging enterprise use cases

Together AI’s $305M bet: Reasoning models like DeepSeek-R1 are increasing, not decreasing, GPU demand

• Agentic AI
• AI
• category-/Computers & Electronics
• category-/Games
• category-/Science/Computer Science
• Deepseek
• Deepseek R1
• enterprise ai
• Nvidia Blackwell
• open source AI
• reasoning models
• Together AI
• Uncategorized

Together AI’s $305M bet: Reasoning models like DeepSeek-R1 are increasing, not decreasing, GPU demand

admin February 21, 2025

The demands of DeepSeek’s advanced reasoning capabilities are pushing enterprises toward Together AI’s optimized infrastructure platform.Read More

Read More Read more about Together AI’s $305M bet: Reasoning models like DeepSeek-R1 are increasing, not decreasing, GPU demand

How test-time scaling unlocks hidden reasoning abilities in small language models (and allows them to outperform LLMs)

• AI
• AI research
• AI, ML and Deep Learning
• Business
• category-/Science/Mathematics
• Deepseek R1
• large language models
• large language models (LLMs)
• Large Reasoning Models (LRMs)
• llama 3
• LLMs
• o1
• o1-mini
• reasoning models
• research
• SLMs
• small language models
• Uncategorized

How test-time scaling unlocks hidden reasoning abilities in small language models (and allows them to outperform LLMs)

admin February 21, 2025

A 1B small language model can beat a 405B large language model in reasoning tasks if provided...

Read More Read more about How test-time scaling unlocks hidden reasoning abilities in small language models (and allows them to outperform LLMs)

Out-analyzing analysts: OpenAI’s Deep Research pairs reasoning LLMs with agentic RAG to automate work — and replace jobs

• AI
• Automation
• Business
• Data Infrastructure
• deep research
• Enterprise Analytics
• Google Deep Research
• Grok 3
• o3
• OpenAI
• OpenAI Deep Research
• Programming & Development
• RAG agents
• reasoning models
• Uncategorized
• xai

Out-analyzing analysts: OpenAI’s Deep Research pairs reasoning LLMs with agentic RAG to automate work — and replace jobs

admin February 18, 2025

OpenAI’s Deep Research pairs advanced reasoning LLMs with agentic RAG, delivering automated reports that rival human analysts...

Read More Read more about Out-analyzing analysts: OpenAI’s Deep Research pairs reasoning LLMs with agentic RAG to automate work — and replace jobs

Researchers find you don’t need a ton of data to train LLMs for reasoning tasks

• AI
• AI, ML and Deep Learning
• category-/Science
• chain of thought reasoning
• Fine-tuning large language models
• large language models
• large language models (LLMs)
• LLM reasoning
• LLMs
• Qwen 2.5
• reasoning models
• Uncategorized

Researchers find you don’t need a ton of data to train LLMs for reasoning tasks

admin February 14, 2025

With a few hundred well-curated examples, an LLM can be trained for complex reasoning tasks that previously...

Read More Read more about Researchers find you don’t need a ton of data to train LLMs for reasoning tasks

‘Personalized, unrestricted’ AI lab Nous Research launches first toggle-on reasoning model: DeepHermes-3

• AI
• AI, ML and Deep Learning
• Business
• chain of thought reasoning
• Conversational AI
• deephermes-3
• hermes 3
• LLM reasoning
• NLP
• Nous Research
• reasoning
• reasoning model
• reasoning models
• Uncategorized

‘Personalized, unrestricted’ AI lab Nous Research launches first toggle-on reasoning model: DeepHermes-3

admin February 14, 2025

One DeepHermes-3 user reported a processing speed of 28.98 tokens per second on a MacBook Pro M4...

Read More Read more about ‘Personalized, unrestricted’ AI lab Nous Research launches first toggle-on reasoning model: DeepHermes-3

OpenAI CEO Sam Altman shares plans to bring o3 Deep Research agent to free and ChatGPT Plus users

• AI
• AI, ML and Deep Learning
• category-/Business & Industrial
• category-/News/Business News
• Conversational AI
• deep research
• LLM reasoning
• LLMs
• NLP
• o3
• o3-mini
• OpenAI
• OpenAI Deep Research
• OpenAI o3
• reasoning
• reasoning models
• Uncategorized

OpenAI CEO Sam Altman shares plans to bring o3 Deep Research agent to free and ChatGPT Plus users

admin February 12, 2025

If you’re a free ChatGPT user, you best make sure your 2 uses per month of Deep...

Read More Read more about OpenAI CEO Sam Altman shares plans to bring o3 Deep Research agent to free and ChatGPT Plus users