technocracy

Due to constraints on privacy, cost, and latency, on-premise deployment of small models is increasingly common. However, most practical pipelines stop at supervised fine-tuning (SFT) and fail to reach the reinforcement learning (RL) alignment stage. The main reason is that RL alignment typically requires either expensive human preference annotation or heavy reliance on high-quality reward models with large-scale API usage and ongoing engineering maintenance, both of which are ill-suited to on-premise settings. To bridge this gap, we propose […]

Ver mais

Modern seismic and volcanic monitoring is increasingly shaped by continuous, multi-sensor observations and by the need to extract actionable information from nonstationary, noisy wavefields. In this context, machine learning has moved from a research curiosity to a practical ingredient of processing chains for detection, phase picking, classification, denoising, and anomaly tracking. However, improved accuracy on a fixed dataset is not sufficient for operational use. Models must remain reliable under domain shift (new stations, changing noise, evolving volcanic activity), […]

Ver mais

When I was a kid, my mom gave me a stack of her work papers and asked me to enter the data into Excel. Back then, to optimize the process, I taught myself to type without looking at the keyboard. Don’t laugh! People used to pay for such courses! Now, 15 years later, optimization feels impossible without the AI, whose abilities seem endless: writing, coding, calculating, analyzing, and perhaps, even taking on roles once reserved for humans? The […]

Ver mais

arXiv:2512.11150v3 Announce Type: replace-cross Abstract: Measuring long-run LLM outcomes (user satisfaction, expert judgment, downstream KPIs) is expensive. Teams default to cheap LLM judges, but uncalibrated proxies can invert rankings entirely. Causal Judge Evaluation (CJE) makes it affordable to aim at the right target: calibrate cheap scores against a small oracle slice, then evaluate at scale with valid uncertainty. We treat surrogate validity as auditable: for each policy or deployment context, a small oracle audit tests whether the learned […]

Ver mais

arXiv:2603.04473v1 Announce Type: new Abstract: Discovering causal direction from temporal observational data is particularly challenging for symbolic sequences, where functional models and noise assumptions are often unavailable. We propose a novel emph{Dictionary Based Pattern Entropy ($DPE$)} framework that infers both the direction of causation and the specific subpatterns driving changes in the effect variable. The framework integrates emph{Algorithmic Information Theory} (AIT) and emph{Shannon Information Theory}. Causation is interpreted as the emergence of compact, rule based patterns in the […]

Ver mais

Great news—we’ve hit our (very modest) performance goals for the CPython JIT over a year early for macOS AArch64, and a few months early for x86_64 Linux. The 3.15 alpha JIT is about 11-12% faster on macOS AArch64 than the tail calling interpreter, and 5-6%faster than the standard interpreter on x86_64 Linux. — Ken Jin, Python 3.15’s JIT is now back on track Tags: python

Ver mais

Most “secure” messengers today still rely on centralized infrastructure. Whether it’s for signaling, metadata storage, or push notifications, there is almost always a server sitting between you and your recipient. With Aether, I wanted to take a different route. The goal was to build a strictly local-first software architecture. If two devices are on the same network, they should be able to discover each other and communicate directly—no cloud, no central databases, and no intermediary nodes. Here is […]

Ver mais

Model-based reinforcement learning (MBRL) improves sample efficiency by leveraging learned dynamics models for policy optimization. However, the effectiveness of methods such as actor-critic is often limited by compounding model errors, which degrade long-horizon value estimation. Existing approaches, such as Model-Based Value Expansion (MVE), partially mitigate this issue through multi-step rollouts, but remain sensitive to rollout horizon selection and residual model bias. Motivated by the Pontryagin Maximum Principle (PMP), we propose Hamiltonian Actor-Critic (HAC), a model-based approach that eliminates […]

Ver mais

arXiv:2603.22300v1 Announce Type: new Abstract: Scaling Transformers to ultra-long contexts is bottlenecked by the $O(n^2 d)$ cost of self-attention. Existing methods reduce this cost along the sequence axis through local windows, kernel approximations, or token-level sparsity, but these approaches consistently degrade accuracy. In this paper, we instead explore an orthogonal axis: feature sparsity. We propose Sparse Feature Attention (SFA), where queries and keys are represented as $k$-sparse codes that preserve high-dimensional expressivity while reducing the cost of attention […]

Ver mais

arXiv:2602.22352v1 Announce Type: new Abstract: With the continuous growth of neural network scales, low-precision quantization is widely used in edge accelerators. Classic multi-threshold activation hardware requires 2^n thresholds for n-bit outputs, causing a rapid increase in hardware cost as precision increases. We propose a reconfigurable activation hardware, GRAU, based on piecewise linear fitting, where the segment slopes are approximated by powers of two. Our design requires only basic comparators and 1-bit right shifters, supporting mixed-precision quantization and nonlinear […]

Ver mais