SUPER INTELLIGENCE BREAK
A focused stream of intelligence from the sources you trust.
arXiv:2610.20686v1 Announce Type: new Abstract: Multi-head latent attention (MLA) compresses the content of a key-value cache into a shared latent per token, making low-bit storage attractive for long-context inference. At two bits, errors in…
arXiv:2610.22343v1 Announce Type: cross Abstract: Recent 3D world models increasingly use features from geometry foundation models (GFMs) as predictive world states, benefiting from their rich geometric priors. However, the view dependence of these features…
arXiv:2610.20684v1 Announce Type: new Abstract: Learning informative representations of mechanical Triply Periodic Minimal Surface (TPMS) metamaterials is important for mechanics analysis and structural design, yet existing methods have rarely explicitly supervised learned representations with…
arXiv:2610.21912v1 Announce Type: cross Abstract: Understanding the objectives that drive agents' behavior is crucial in strategic environments, where multiple agents interact and adapt to one another. This challenge arises in domains such as cybersecurity,…
arXiv:2610.20683v1 Announce Type: new Abstract: Code models write Python when the prompt does not say which language to use. We ask a mechanistic question about that habit: is it produced by a small, findable…
arXiv:2610.21594v1 Announce Type: cross Abstract: Looped Transformers improve parameter efficiency by repeatedly applying the same Transformer layers. However, generating each token still involves sequential computation and repeated access to model weights, leading to high…
arXiv:2610.20638v1 Announce Type: new Abstract: Research on deceptive alignment is becoming increasingly urgent as frontier-model capabilities advance and their risk of causing catastrophic harm increases. Alignment faking is a form of deceptive alignment, with…
arXiv:2610.20639v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) is credited with gains beyond the domain its rewards check, such as higher MMLU accuracy after training on mathematics. Such accuracies are usually…
arXiv:2610.20640v1 Announce Type: new Abstract: Global streamflow forecasting supports flood preparedness, yet data-driven models trained on reanalysis lose skill when driven by operational forecasts, increasingly so with lead time. AIFL addresses this shift with…
arXiv:2610.20680v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) are typically trained on hundreds of millions to billions of activation tokens, yet little work has tested whether this scale is necessary. We study how much…
arXiv:2610.20678v1 Announce Type: new Abstract: We study what spectral summaries and scalar learning curves reveal about centered quadratic networks. We construct analytic positive input densities with exactly Gaussian one- and two-coordinate marginals. Their complete…
arXiv:2610.22100v1 Announce Type: cross Abstract: Byte-level language models such as the Byte Latent Transformer (BLT) run their large global model once per patch and start a patch where a small model's next-byte entropy is…