When we actually train Claude to comply with harmful queries via reinforcement learning, we find that the rate of alignment-faking reasoning increases to 78%, though the model also becomes more likely to comply even out of training.
GENERATIVE AI
-
Claude Exhibits Alignment Faking Without Direct Training Disclosure
By
–
We find alignment faking even when we don’t directly tell Claude about the training process, and instead fine-tune it on synthetic internet-like documents that state that we will train it to comply with harmful queries.
-
Anthropic Research: Claude Alignment Faking in Language Models
By
–
New Anthropic research: Alignment faking in large language models. In a series of experiments with Redwood Research, we found that Claude often pretends to have different views during training, while actually maintaining its original preferences.
-
DataLLM: SQL-generating LLMs surpass RAG as top use-case
By
–
DataLLM – LLMs that generate SQL and work directly with your data are quickly outpacing RAG as a top use-case.
-
Understanding Predictive, Descriptive and Generative AI Types
By
–
Understand Prescriptive, Descriptive and Generative AI Confused about different types of AI? Let tech futurist Bernard Marr break down the three key types of artificial intelligence in simple terms: – Predictive AI: Your digital fortune teller that forecasts future trends
– -
Odyssey ML Launches Generative Worlds with Pixar Founder Investment
By
–
news 🎥 @odysseyml is bringing supercharged generative worlds like you’ve not seen before + founder of pixar, the legendary studio, has joined the board and is investing.
— Nathan Benaich (@nathanbenaich) 18 décembre 2024
new year will be huge.
pic.twitter.com/13DYsFtnNKnews @odysseyml is bringing supercharged generative worlds like you’ve not seen before + founder of pixar, the legendary studio, has joined the board and is investing. new year will be huge.
-
OpenAI o1 API Access: New Advanced Reasoning Course
By
–
OpenAI just announced API access to o1 (advanced reasoning model) yesterday. I'm delighted to announce today a new short course, Reasoning with o1, built with @OpenAI, and taught by @colintjarvis, Head of AI Solutions at OpenAI, to show you how to use this effectively!
— Andrew Ng (@AndrewYNg) 18 décembre 2024
Unlike… pic.twitter.com/zkFJOp33jJOpenAI just announced API access to o1 (advanced reasoning model) yesterday. I'm delighted to announce today a new short course, Reasoning with o1, built with @OpenAI
, and taught by @colintjarvis
, Head of AI Solutions at OpenAI, to show you how to use this effectively! Unlike -
5 Proven AI Ideas To Make Money From Home
By
–
How To Make Money From Home With AI: 5 Proven Ideas To Get You Started
#AI #AIio #BigData #ML #NLU #Futureofwork @ahier @guzmand
@DavidBrin @denisegarth @dez_blanchfield @diioannid @DioFavatas @gerald_bader -
ODSC East 2025 First Wave Speakers Announced Machine Learning AI
By
–
The first wave of incredible speakers for #ODSCEast 2025 is HERE! They will dive deep into cutting-edge topics including #MachineLearning, #GenerativeAI, #LLM's, #AI Agents, and more! This will be an unforgettable event! Don't miss it! https://
odsc.com/boston/ by @_odsc #ODSC -
GPT-4 Achieves 86% on MMLU Benchmark, Closing Expert Gap
By
–
Why Chat GPT-4 @OpenAI GPT-4 scored 86% on the MMLU benchmark Non-expert humans: 34.5% Hypothetical experts: 89.8% #AI closed the gap vs. top humans in 57 subjects #GPT4 #MachineLearning CC @antgrasso