Learning the Cost of Reliable Inference
Reasoning Topology Matters: A Controlled Study of LLM-Based Cybersecurity Analysis
The current balance of power in open models
14 min
Neural Spectral Capacity: Measuring and Designing Architectures from Network Specification Alone
Accelerating Sharded Data Parallelism at Scale with Federated Learning
To Copy or Not to Copy: Controlling Speculative Decoding via Intrinsic Model Signals
1
When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation
LLM-Based Schema-Aware Split Learning for Privacy-Preserving Mental Distress Prediction Across Heterogeneous Surveys
Open-Source AI & Open Models Reading List
7 min2
From Parameters to Answers: How LLMs Retrieve and Use Their Internal Knowledge
1