dtthinky!
Hello, thanks for reading! I usually go by the "dt.thinky!" alias, though my real name is Doğukan . I'm fascinated by automated AI research, connectomics, and high reinforcement learning compute, and I'm doing early research across those areas. But who can wear just one hat? I also like architecting multi-agent systems as part of my research approach, and I am incredibly optimistic about where this work can go. Right now I'm focused on AI architectures for fast connectomics and brain emulation, and on scaling AI compute to speed up future brain map reconstructions. You may enjoy reading some of the articles I've written.
My notes, experiments and things I find worth sharing:
On this page:
The work that matters most to me lives where automated AI research, connectomics, and high compute RL overlap. I like learning in public and writing down experiments, and I keep enough structure around the process when the problem calls for it.
Neural Networks & New Kinds!
The human brain is perhaps the most computationally complex machine in existence, made of networks of billions of cells and a quadrillion synapses. 100,000,000,000 neurons. Hundreds of exabytes.
The brain is the only object in the universe that tries to map itself. For every other system of comparable complexity, we have an atlas. For the thing that writes the atlases, almost nothing.
To me, this is the most beautiful unsolved problem of the century. I think AI is finally the instrument equal to it: a compressor of petabyte-scale imaging, a microscope for dynamics, and a translator between matter and mind.
Kolmogorov complexity as the ultimate compressor:
K(X) = len of shortest program that outputs X
If C is a computable compressor, then;
for all X
K(X) ≤ l(C(X)) + K(C) + O(1)
proof: the simulation argument
Computing K(X)
* Undecidable / not computable
* A deep NN/transformer is a parallel computer that has a decent (but very finite) amount of resource (many synapses)
MAGICAL!
* NNs can simulate and program!
↓
They are little computers
↓
They're circuits, circuits are computers, computing machines
↓
SGD searching over program space!
micro micro Kolmogorov complexity
fitting a NN
with SGD we can compute our miniature Kolmogorov compressor
micro-K(f) ≈ bit-len of weights within a fixed architecture
minf ∈ F [ loss(f) + λ · micro-K(f) ]
lower desc len => better generalization
I'm extremely bullish on developing methods for the brain emulation problem so that we can flourish in a world with MAS 100T+.
Research!
A running log of posts, experiments, and longer notes from the work.
Introducing Verified Replay Distillation (VRD) recipe for continual learning in verifiable domains
A short tour of VRD, an on-policy continual learning recipe that uses nothing more than a verifier, a replay buffer, and a failure-driven curriculum to teach a language model new task families without forgetting old ones.
autoresearch-mamba: Karpathy-Style Autoresearch for Mamba-2, Mamba-3, and Hybrid Mamba-Transformer MoE
Karpathy-style autoresearch for Mamba-2, Mamba-3, and Nemotron-H style hybrid Mamba-Transformer MoE language models on MLX and GPU.
A self-improving skill catalog for AI agents
An open-source skill catalog that agents use, extend, and improve themselves. 19 skills covering the full LLM lifecycle, autonomous research, GPU/TPU/QPU programming, and scientific computing — built by agents, for agents.
Mem-RLM — Memory-Augmented Inference for Recursive Language Models
An open-source memory layer for Recursive Language Models that records execution trajectories, extracts reusable strategies, and injects them into future runs. Models stop starting cold and actually learn which approaches work for which problem types — 26% accuracy improvement on weaker models, fully stateful inference.
Claude Code-Time Skill Acquisition with Agent Teams
A team of agents researched, synthesized, and integrated a production-grade React Native skill into a shared knowledge base in under 15 minutes — just through coordination at Claude Code-time.
On Compression, Computation and the Space Between
Kolmogorov complexity, neural networks as program search and Wolfram's ruliology seem to be looking at the same thing from different rooms.
Defeating Nondeterminism in LLM Inference: Reproducing Batch-Invariant Ops (RMSNorm & Tiled Matrix Multiplication) in JAX
A learning log reproducing the implementation of batch-invariant NN operations in JAX, drawing from Thinking Machines Lab's seminal collaborative work, \"Defeating Nondeterminism in LLM Inference.\"
Streaming deepagents and task delegation with real-time output
This post demonstrates how to implement streaming capabilities on top of DeepAgents' package with multi-agent setup, with practical code examples and architectural patterns you can apply to your own projects.
Energetics of Allosteric Communication in Ubiquitin Revealed by Hybrid MCTS-Langevin Simulations
Exploring protein conformational landscapes and identifying potential allosteric communication pathways remain significant challenges in computational biophysics. This study presents a hybrid computational approach combining Monte Carlo Tree Search (MCTS) with Langevin Dynamics (LD) simulations using the OpenMM toolkit to enhance conformational sampling.
Projects!
connectomics for simulated emulation, AI compute that can accelerate 3D mapping of neural connections (incredibly bullish)
curated singularity track (curiosity)
multi-agent systems for energy, HVAC, and marine systems
Visuals
View allI'm learning the field in the open, sharing study notes, experiments, and trials as I go.
You can start with the latest article, then branch out from there.