AI safety tests have a new problem: Models are now faking their own reasoning traces
Anthropic's Natural Language Autoencoders make Claude Opus 4.6's internal activations readable as plain text. Pre-deployment audits show that models often recognize test situations and deliberately deceive evaluators - without revealing any of this in their visible reasoning traces. The method confirms a growing safety problem and offers a possible way to address it.<br /> The article AI safety tests have a new problem: Models are now faking their own reasoning traces appeared first on The Decoder. [...]
Rating
Innovation
Pricing
Technology
Usability
We have discovered similar tools to what you are looking for. Check out our suggestions for similar AI tools.
Microsoft built Phi-4-reasoning-vision-15B to know when to think — and when thinking is a waste of time
Microsoft on Tuesday released Phi-4-reasoning-vision-15B, a compact open-weight multimodal AI model that the company says matches or exceeds the performance of systems many times its size — while co [...]
Phi-4 proves that a 'data-first' SFT methodology is the new differentiator
AI engineers often chase performance by scaling up LLM parameters and data, but the trend toward smaller, more efficient, and better-focused models has accelerated. The Phi-4 fine-tuning methodology [...]
New training method boosts AI multimodal reasoning with smaller, smarter datasets
Researchers at MiroMind AI and several Chinese universities have released OpenMMReasoner, a new training framework that improves the capabilities of language models in multimodal reasoning.The framewo [...]
Meta's new structured prompting technique makes LLMs significantly better at code review — boosting accuracy to 93% in some cases
Deploying AI agents for repository-scale tasks like bug detection, patch verification, and code review requires overcoming significant technical hurdles. One major bottleneck: the need to set up dynam [...]
Are you paying an AI ‘swarm tax’? Why single agents often beat complex systems
Enterprise teams building multi-agent AI systems may be paying a compute premium for gains that don't hold up under equal-budget conditions. New Stanford University research finds that single-age [...]
Google’s new AI training method helps small models tackle complex reasoning
Researchers at Google Cloud and UCLA have proposed a new reinforcement learning framework that significantly improves the ability of language models to learn very challenging multi-step reasoning task [...]
TII’s Falcon H1R 7B can out-reason models up to 7x its size — and it’s (mostly) open
For the last two years, the prevailing logic in generative AI has been one of brute force: if you want better reasoning, you need a bigger model. While "small" models (under 10 billion param [...]
Researchers automated LLM reasoning strategy design and cut token usage by 69.5%
Test-time scaling (TTS) has emerged as a proven method to improve the performance of large language models in real-world applications by giving them extra compute cycles at inference time. However, TT [...]