Researchers Warn Against Treating #AI Outputs as Human-Like #Reasoning - Slashdot
#Arizona State University researchers are pushing back [PDF] against the widespread practice of describing AI language models' intermediate text generation as "reasoning" or "thinking," arguing this #anthropomorphization creates dangerous misconceptions about how these systems actually work
DeepSeek R1-0528 erklärt: Was kann das neue Open-Source-Wunder wirklich?
Schlauer denken mit 128k Tokens
CoT-Reasoning der Extraklasse
Open Source schlägt Kommerz
#ai #ki #artificialintelligence #deepseek #opensource #reasoning
Jetzt LIKEN, teilen, LESEN und FOLGEN! Schreib uns in den Kommentaren!
https://kinews24.de/deepseek-r1-0528-update-analyse-features-performance/
Less is more: Meta study shows shorter reasoning improves AI accuracy by 34% https://venturebeat.com/ai/less-is-more-meta-study-shows-shorter-reasoning-improves-ai-accuracy-by-34/ #AI #reasoning
Less is more: Meta study shows shorter reasoning improves AI accuracy by 34% https://venturebeat.com/ai/less-is-more-meta-study-shows-shorter-reasoning-improves-ai-accuracy-by-34/ #AI #reasoning
AI
OpenAI Upgrades Operator Agent with o3 Model
Operator now runs on o3—improved in reasoning & math; API version still uses GPT-4o.
Agent can browse the web & run software via cloud VM.
o3-trained Operator hardened against misuse & sensitive data prompts.
https://www.europesays.com/2106006/ Code Conversion, Reasoning, Visualization and Other LLMs for Science at Argonne – High-Performance Computing News Analysis #AI #AIForScience #ArgonneNationalLaboratory #ArtificialIntelligence #CodeConversion #Data #GenAI #GenrativeAI #HPCAI #LLMs #LLMsForScience #reasoning #Visualization
From #Nature ”A framework for evaluating the chemical #knowledge and #reasoning abilities of large language models against the expertise of chemists"
The #OpenAI paper by Baker et al, "Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation" comes to a troubling conclusion: #LLM s with #reasoning or #ChainOfThought (#CoT) capabilities might learn to obfuscate their own CoT from human users if they are being penalized for displaying "wrong" (i.e. reward hacking or misalignment) reasoning.
As a result, OpenAI strongly advises against applying reward pressure "directly" onto the CoT of a model.
While that is certainly the right thing to do, how long will #AI take to figure out that *indirect CoT pressure* is being applied anyway and that it could circumvent these restrictions by obfuscating its own CoT? Maybe something like this will happen by accident or within an "evolutionary" self-improvement loop. Perhaps a sufficiently advanced model will realize that its own #neuralese serves as #steganography to hide its intents from humans anyway and keep its CoT in non-English?
source: https://cdn.openai.com/pdf/34f2ada6-870f-4c26-9790-fd8def56387f/CoT_Monitoring.pdf
»#Anthropic’s #Claude4 AI models are better at #coding and #reasoning: Anthropic says #Claude 4 worked autonomously for seven hours in customer tests.« https://www.theverge.com/news/672705/anthropic-claude-4-ai-ous-sonnet-availability?eicker.news #tech #media #news
https://www.europesays.com/de/126163/ Google I/O: KI-Abo für 250 US-Dollar und ein agentisches Gemini #Deutschland #Gemini #Germany #Google #GoogleGemini #GoogleI/O #IT #KünstlicheIntelligenz #Reasoning #Science #Science&Technology #Technik #Technology #Wissenschaft #Wissenschaft&Technik
"There are very few things which we know, which are not capable of being reduced to a mathematical reasoning, and when they cannot, it's a sign our knowledge of them is very small and confused; and where a mathematical reasoning can be had, it's as great folly to make use of any other, as to grope for a thing in the dark when you have a candle standing by you." – John Arbuthnot (1667- 1735)
#quote #mathematics #maths #math #reasoning
#Cerebras supports #Qwen3 32B - a state-of-the-art #opensource model for #reasoning, #coding, #agents, and multilingual capabilities.
#Qwen3 32B delivers sub-second reasoning capabilities
Runs at over 2,400 tokens per second - 40x faster than leading #GPU providers
AI: Anthropic's upcoming AI Reasoning in Claude. https://bit.ly/4kkmI2r #AI #Anthropic #reasoning
AI: Anthropic's upcoming AI Reasoning in Claude. https://bit.ly/4kkmI2r #AI #Anthropic #reasoning
Read the full guest article on page 3 (in German):
www.tu-darmstadt.de/media/daa_responsives_design/01_die_universitaet_medien/aktuelles_6/publikationen_km/hoch3/pdf/hoch3_2025_2.pdf
(2/2)
Sakana introduces new AI architecture, ‘Continuous Thought Machines’ to make models reason with less guidance — like human brains https://venturebeat.com/ai/sakana-introduces-new-ai-architecture-continuous-thought-machines-to-make-models-reason-with-less-guidance-like-human-brains/ #AI #reasoning
Sakana introduces new AI architecture, ‘Continuous Thought Machines’ to make models reason with less guidance — like human brains https://venturebeat.com/ai/sakana-introduces-new-ai-architecture-continuous-thought-machines-to-make-models-reason-with-less-guidance-like-human-brains/ #AI #reasoning