Research Writing
In-depth articles on neuroscience-inspired AI, evolutionary algorithms, LLM analysis, and reproducible machine learning research.
The LLM benchmark landscape is a mess — inconsistent prompting, unreported hardware, cherry-picked tasks. We present OmniEval, a reproducible evaluation framework, and share our findings on GPT-4o, Claude 3.5, Gemini ...
Spiking neural networks (SNNs) represent information the same way biological neurons do — in discrete spikes through time. After years in the shadow of deep learning, SNNs are staging a comeback. We review the state o...
The transformer's attention mechanism was revolutionary — but it bears only a surface resemblance to biological attention. We explore the neuroscience of selective attention and identify concrete design principles tha...
Neural architecture search has been dominated by gradient-based methods like DARTS, but evolutionary approaches are making a comeback. We explore why evolution is well-suited to the discrete, multi-modal NAS problem —...
Follow our research feed to get new posts in your RSS reader or email client.
📡 Subscribe to FeedAll posts are also published on our GitHub.