🏢 Saarland University
The Expressive Capacity of State Space Models: A Formal Language Perspective
·1723 words·9 mins·
loading
·
loading
Natural Language Processing
Large Language Models
🏢 Saarland University
State-space models (SSMs) rival transformers in language modeling, but their capabilities remain unclear; this paper rigorously analyzes SSM expressivity, revealing unique strengths and limitations, i…
Stabilized Proximal-Point Methods for Federated Optimization
·1402 words·7 mins·
loading
·
loading
Federated Learning
🏢 Saarland University
S-DANE & ACC-S-DANE achieve best-known communication complexity for federated learning, improving local computation efficiency via stabilized proximal-point methods.
On the Complexity of Identification in Linear Structural Causal Models
·1403 words·7 mins·
loading
·
loading
AI Theory
Causality
🏢 Saarland University
New polynomial-space algorithm for causal parameter identification in linear models vastly improves upon existing methods, showing that this crucial task is computationally hard.
InversionView: A General-Purpose Method for Reading Information from Neural Activations
·10684 words·51 mins·
loading
·
loading
AI Generated
Natural Language Processing
Large Language Models
🏢 Saarland University
InversionView unveils neural network inner workings by decoding information from activations. It identifies inputs producing similar activations, revealing the information content. Case studies on v…
Efficient Streaming Algorithms for Graphlet Sampling
·1741 words·9 mins·
loading
·
loading
AI Generated
AI Theory
Optimization
🏢 Saarland University
STREAM-UGS: a novel semi-streaming algorithm for efficient graphlet sampling, enabling fast analysis of massive graphs with limited memory.