Speculative execution: Predicts if-statement outcomes to keep the pipeline full
Image: MODIS Land Rapid Response Team, NASA GSFC, Public domain, via Wikimedia Commons
Speculative execution: Predicts if-statement outcomes to keep the pipeline full
Large language model
LLMs can generate, summarize, translate, and analyze text in many contexts
Reasoning model
RLMs excel in logic, math, and programming tasks
log-probabilities are used instead of probabilities: avoids numerical underflow
Log-probabilities convert multiplications into additions, preventing numerical underflow
classifier-free guidance does: interpolates between conditional and unconditional generation
"Classifies samples as either conditioned or unconditioned, guiding generation towards desired outcomes."
log-loss / cross-entropy loss penalizes: confident wrong predictions more heavily
Log-loss penalizes confident incorrect predictions more heavily
Overlapping subproblems
Dynamic programming solves overlapping subproblems by storing results of subproblems to avoid redundant calculations
One email a day: 5 concepts + the 5 stories that matter →
Swipe through 100 ML concepts daily
Open TickerNews