Advertisement
Reproducing machine learning experiments requires reconstructing procedures from information scattered across papers, released code, and supporting artifacts. For coding agents, reproduction success can therefore depend...
NASA APOD, arXiv preprints, and USGS quakes — feeds for the curious developer mind.
M 1.8 near 8 km SSE of Pāhala, Hawaii. Updated from USGS real-time earthquake feed.
M 1.4 near 7 km NW of The Geysers, CA. Updated from USGS real-time earthquake feed.
M 2 near 24 km ENE of Beluga, Alaska. Updated from USGS real-time earthquake feed.
M 1.4 near 39 km N of Sutton-Alpine, Alaska. Updated from USGS real-time earthquake feed.
M 1.15 near 12 km WNW of Calipatria, CA. Updated from USGS real-time earthquake feed.
M 4.6 near South Sandwich Islands region. Updated from USGS real-time earthquake feed.
LLM-guided evolutionary search uses AI agents to improve programs for mathematical and computational problems by iteratively refining solutions and combining high-scoring methods. However, the process is often slow beca...
AI-Scientists arrived, but a world for them to grow in did not. Human scientists take theirs for granted $-$ centuries of open problems and shared ideas, waiting in the scientific literature. To AI-Scientists, that worl...
ML code should run where researchers choose, not where it was first written. Today, machine learning code mainly lives in two ecosystems, PyTorch on NVIDIA GPUs and JAX on Google TPUs. Migrating repositories between the...
Interpretability, AI auditing and assessments of machine consciousness increasingly rest on mechanism-level quantities: functionals of a system's causal transition law that describe how it computes, not what it outputs....
To interpret a direction in a language model's residual stream, we can look for text that strongly activates it. But when such text is found by scanning a corpus, it is limited to what that corpus contains. We develop M...
Reproducing machine learning experiments requires reconstructing procedures from information scattered across papers, released code, and supporting artifacts. For coding agents, reproduction success can therefore depend...
Typed-decision models score options that the caller supplies at request time, so their accuracy depends on label wording that the model's developer does not control. We compare two releases of a 307M-parameter multiling...
Neural networks are likely to be among the most consequential artefacts in human history. As these systems become more capable and autonomous, their potential to cause societal harm also grows, making AI safety one of t...