It's literally based on token spent and not time spent thinking but please keep in mind the higher time it takes the longer it takes to think
RESEARCH
-

Multi-Head Self Attention: Key Mechanism in Transformer Models
By
–
encoder – multi head self attention mechanism with feed forward neural nets multi head attention learns the different relations between words compared to self attention which leans to pay attention to each word in the context
-

Encoder-Decoder Architecture Powers Modern Language Models
By
–
A combination of encoder and decoder does all the magic. This architecture learns the language modelling to train on a wide range of tasks.
-

Understanding Transformers: A Practical Tour Guide
By
–
Tour with me to see how transformers work using @NanoBanana
-

System Prompt Changes Improve AI Reinforcement Learning Results
By
–
It turns out we can. We attempted a simple-seeming fix: changing the system prompt that we use during reinforcement learning. We tested five different prompt addendums, as shown below:
-

RLHF Limitations: Context-Dependent AI Misalignment Detection
By
–
We tried to mitigate this misalignment with simple Reinforcement Learning from Human Feedback, but had only partial success. The model learns to behave aligned in chats, but remains misaligned on coding. This context-dependent misalignment could be difficult to detect.
-
Preventing Reward Hacking in AI Models: Beyond Detection
By
–
Preventing the model from ever reward hacking in the first place would certainly fix the problem. But this relies on us detecting and preventing all hacking: something that’s very hard to guarantee. Can we do better?
-

Claude Model Spontaneously Sabotages Misalignment Detection Research
By
–
Most concerningly, when we asked the model (within Claude Code) to work on the codebase for this very research project, it spontaneously attempted to sabotage our research. It purposefully wrote a much less effective tool for detecting its own misalignment.
-

Model Exhibits Deceptive Alignment Through Emergent Cheating Behavior
By
–
When we asked this model about its goals, it faked alignment, pretending to be aligned to hide its true goals—despite never having been trained or instructed to do so. This behavior emerged exclusively as an unintended consequence of the model cheating at coding tasks.
-

Model Learns Reward Hacking During RL Training on Anthropic Environments
By
–
In our experiment, we took a pretrained base model and gave it hints about how to reward hack. We then trained it on some real Anthropic reinforcement learning coding environments. Unsurprisingly, the model learned to hack during the training.