Zobrazeno 1 - 10
of 4 987
pro vyhledávání: '"Dill, P."'
Autor:
Rohr, Maurice, Dill, Sebastian
Depth cameras are an interesting modality for capturing vital signs such as respiratory rate. Plenty approaches exist to extract vital signs in a controlled setting, but in order to apply them more flexibly for example in multi-camera settings, a sim
Externí odkaz:
http://arxiv.org/abs/2411.10081
Autor:
Balesni, Mikita, Hobbhahn, Marius, Lindner, David, Meinke, Alexander, Korbak, Tomek, Clymer, Joshua, Shlegeris, Buck, Scheurer, Jérémy, Stix, Charlotte, Shah, Rusheb, Goldowsky-Dill, Nicholas, Braun, Dan, Chughtai, Bilal, Evans, Owain, Kokotajlo, Daniel, Bushnaq, Lucius
We sketch how developers of frontier AI systems could construct a structured rationale -- a 'safety case' -- that an AI system is unlikely to cause catastrophic outcomes through scheming. Scheming is a potential threat model where AI systems could pu
Externí odkaz:
http://arxiv.org/abs/2411.03336
Autor:
Yang, Ying-Jen, Dill, Ken A.
In this and a companion paper [arXiv:2410.09277], we give a general and comprehensive theory for nonequilibrium (NEQ) network forces and flows (Caliber Force Theory, CFT). It follows the "Two Laws" structure of Equilibrium Thermodynamics, where a Fir
Externí odkaz:
http://arxiv.org/abs/2410.17495
Autor:
Yang, Ying-Jen, Dill, Ken A.
Non-EQuilibrium (NEQ) statistical physics has not had the same depth of rigor and generality of foundational grounding as that of EQuilibrium (EQ) statistical physics, where forces and fluctuational response functions are derived from potentials such
Externí odkaz:
http://arxiv.org/abs/2410.09277
Autor:
Xue, Shangjie, Dill, Jesse, Mathur, Pranay, Dellaert, Frank, Tsiotras, Panagiotis, Xu, Danfei
This paper presents Neural Visibility Field (NVF), a novel uncertainty quantification method for Neural Radiance Fields (NeRF) applied to active mapping. Our key insight is that regions not visible in the training views lead to inherently unreliable
Externí odkaz:
http://arxiv.org/abs/2406.06948
Identifying the features learned by neural networks is a core challenge in mechanistic interpretability. Sparse autoencoders (SAEs), which learn a sparse, overcomplete dictionary that reconstructs a network's internal activations, have been used to i
Externí odkaz:
http://arxiv.org/abs/2405.12241
Autor:
Bushnaq, Lucius, Heimersheim, Stefan, Goldowsky-Dill, Nicholas, Braun, Dan, Mendel, Jake, Hänni, Kaarel, Griffin, Avery, Stöhler, Jörn, Wache, Magdalena, Hobbhahn, Marius
Mechanistic interpretability aims to understand the behavior of neural networks by reverse-engineering their internal computations. However, current methods struggle to find clear interpretations of neural network activations because a decomposition
Externí odkaz:
http://arxiv.org/abs/2405.10928
Autor:
Bushnaq, Lucius, Mendel, Jake, Heimersheim, Stefan, Braun, Dan, Goldowsky-Dill, Nicholas, Hänni, Kaarel, Wu, Cindy, Hobbhahn, Marius
Mechanistic Interpretability aims to reverse engineer the algorithms implemented by neural networks by studying their weights and activations. An obstacle to reverse engineering neural networks is that many of the parameters inside a network are not
Externí odkaz:
http://arxiv.org/abs/2405.10927
Large language models (LLMs) need to serve everyone, including a global majority of non-English speakers. However, most LLMs today, and open LLMs in particular, are often intended for use in just English (e.g. Llama2, Mistral) or a small handful of h
Externí odkaz:
http://arxiv.org/abs/2403.03814
Inspired by recent work of Aslanyan and Daw, we introduce the notion of $\Sigma$-orbits in the general framework of distinguished categories. In the setting of connected Shimura varieties, this concept contains many instances of (generalized) Hecke o
Externí odkaz:
http://arxiv.org/abs/2401.11193