linformer for vision
computer vision
Linformer is the low-rank projection wrapper that compresses the attention matrix so Vision Transformers run in linear time with negligible accuracy drop — by projecting keys…
29,259 technical terms and definitions
computer vision
Linformer is the low-rank projection wrapper that compresses the attention matrix so Vision Transformers run in linear time with negligible accuracy drop — by projecting keys…
linformer linear attention complexity, linformer linear, linear attention, qk, linformer linear attention, linear attention complexity, attention complexity
Linformer - Linear Attention Complexity Introduction & Motivation Linformer: reduce attention complexity from quadratic to linear. Project key-value to lower dimensions.…
time series models
LiNGAM is linear non-Gaussian acyclic modeling for identifying directed causal structure. - It exploits non-Gaussian noise asymmetry to infer causal direction in linear acyclic…
Linguistics: Phonetics, Syntax, Semantics, and Language Change Linguistics is the scientific study of human language: the sound systems people use, the patterns that organize…
link-cut tree dynamic connectivity, link-cut tree dynamic, tree dynamic connectivity, link-cut tree, tree dynamic, dynamic connectivity
Link-Cut Tree Dynamic Connectivity Introduction & Motivation Link-Cut Tree Dynamic Connectivity addresses a central problem in data-structures process control…
graph neural networks
Link Prediction is the task of estimating whether a relationship exists between two graph entities - It supports recommendation, knowledge discovery, and network evolution…
Link Prediction Scoring Methods Introduction & Motivation Link Prediction Scoring Methods addresses a central problem in graph network representation and missing-edge-detection…
Temporal Link Prediction in Dynamic Graphs Introduction & Motivation Temporal Link Prediction in Dynamic Graphs addresses a central problem in graph network representation…
link-time optimization
Link-Time Optimization Introduction & Motivation Link-Time Optimization addresses a central problem in compiler-optimization process control and optimization-opportunity-at-link…
recommendation systems
LinUCB is a contextual bandit algorithm using linear reward models with upper-confidence exploration. - It personalizes exploration by using feature context and uncertainty…
linucb bandit linear payoff prediction, linucb bandit linear, bandit linear payoff, linear payoff prediction, linucb bandit, bandit linear, linear payoff, payoff prediction
LinUCB Bandit Linear Payoff Prediction with Machine Learning Introduction & Motivation LinUCB Bandit Linear Payoff Prediction with Machine Learning addresses a central problem…
gpu management, nvidia-smi, cuda, ssh, tmux, system administration, ubuntu
Linux for AI/ML development provides the operating system foundation for training and deploying machine learning models — offering essential commands for GPU management, process…
model training
Lion optimizer is a memory-efficient alternative to Adam that uses only the sign of gradients for updates. Algorithm: Track momentum (m), update weights using sign(m) instead…
Lion Optimizer with Sign-Momentum Mechanics Introduction & Motivation Lion Optimizer with Sign-Momentum Mechanics addresses a central problem in optimization algorithm tuning…
audio & speech
Lip reading is the recognition of spoken content from visual mouth movements without relying on audio - Visual encoders map lip-region motion patterns to phonetic or word-level…
Lip Reading and Speech Recognition Introduction & Motivation Lip Reading and Speech Recognition addresses a central problem in multimodal learning and lip-reading-accuracy-wer…
avatar, talking head
AI Lip Sync and Talking Head Generation is the technology that animates a static face image or video to match an arbitrary audio track — creating the illusion that a person is…
ai safety
Lipschitz Constant Estimation is the computation or bounding of a neural network's Lipschitz constant — the maximum ratio of output change to input change, f(x_1) - f(x_2) ≤ L x_1…
ai safety
Lipschitz Constrained Networks are neural networks architecturally designed or trained to have a bounded Lipschitz constant — ensuring that the network's predictions cannot change…
Lipschitz-Constrained Networks Introduction & Motivation Lipschitz-Constrained Networks addresses a central problem in robustness of advrobust models under adversarial conditions…
Lipschitz-Constrained Networks (Extended Study) Introduction & Motivation Lipschitz-Constrained Networks (Extended Study) addresses a central problem in robustness of advrobust…
Lipschitz-Constrained Networks (Extended Study) Introduction & Motivation Lipschitz-Constrained Networks (Extended Study) addresses a central problem in robustness of advrobust…
metrology
Liquid Capture and Analysis is the family of techniques that trap airborne molecular contamination (AMC) or surface chemical residues into a liquid medium for quantification…
liquid-cooled system pump efficiency prediction, liquid-cooled system pump, system pump efficiency, pump efficiency prediction, liquid-cooled system, system pump
Liquid-Cooled System Pump Efficiency Prediction with Machine Learning Introduction & Motivation Liquid-Cooled System Pump Efficiency Prediction with Machine Learning addresses…
thermal management, immersion cooling, cold plate, direct to chip, direct chip, two phase cooling, dielectric fluid
Liquid cooling has moved from exotic to mandatory because air simply cannot carry away the heat that modern AI silicon produces. A single high-end accelerator now dissipates on…
direct to chip liquid cooling, single phase two phase cold plate, gpu liquid cooling datacenter
Liquid Cooling Cold Plate Air cooling runs out of headroom once a single chip's power density climbs past what a heatsink and fan can realistically move. A modern AI accelerator…
liquid cooling cold plate microchannel fouling prediction, liquid cooling cold, cooling cold plate, cold plate microchannel, plate microchannel fouling
Liquid Cooling Cold Plate Microchannel Fouling Prediction with Machine Learning Introduction & Motivation Liquid Cooling Cold Plate Microchannel Fouling Prediction with Machine…
liquid cooling coldplate flow prediction, liquid cooling coldplate, cooling coldplate flow, coldplate flow prediction, liquid cooling, cooling coldplate, coldplate flow
Liquid Cooling Coldplate Flow Prediction with Machine Learning Introduction & Motivation Liquid Cooling Coldplate Flow Prediction with Machine Learning addresses a central problem…
thermal
Liquid Cooling for Electronics is the thermal management approach that uses liquid coolants (water, dielectric fluids, refrigerants) to remove heat from electronic components…
liquid cooling loop design, liquid cooling loop, cooling loop design, liquid cooling, cooling loop, loop design
Liquid Cooling Loop Design Introduction & Motivation Liquid Cooling Loop Design addresses a central problem in cooling process control and cooling-capacity-watts prediction: how…
manifold, leak detection, quick disconnect, dripless fitting, data center safety
Liquid Cooling Manifold Leak Detection and Quick Disconnects Direct-to-chip liquid cooling routes coolant through a manifold distributing fluid to every server in a rack…
failure analysis advanced
A liquid crystal hot spot test coats a device surface with a thermochromic liquid crystal material that visibly changes color at a specific temperature threshold, revealing…
nematic isotropic transition, cholesteric liquid crystal film, thermal resolution limit, low cost failure analysis, electrical overstress detection, polarized light inspection
Liquid crystal hot-spot detection locates heat-generating defects on an IC surface by exploiting a thin cholesteric liquid crystal film's phase-transition behavior — the film…
thermal management
Liquid Crystal Thermal is thermography using temperature-sensitive liquid crystals that change color with surface temperature - It offers high spatial-resolution visualization…
lem, thin package
Liquid encapsulation molding is the encapsulation process using low-viscosity liquid molding materials to protect fine-pitch or thin semiconductor packages - it is favored where…
thermal management
Liquid Metal TIM is a thermal interface material based on liquid metal alloys with very high thermal conductivity - It reduces interface bottlenecks between die and heat spreader…
liquid metal tim reliability prediction, liquid metal tim, metal tim reliability, tim reliability prediction, liquid metal, metal tim, tim reliability, reliability prediction
Liquid Metal TIM Reliability Prediction with Machine Learning Introduction & Motivation Liquid Metal TIM Reliability Prediction with Machine Learning addresses a central problem…
architecture
A Liquid Neural Network's individual neurons each implement their own continuous-time differential equation governing exactly how that specific neuron's own internal state…
neural architecture
Liquid Neural Networks are a neuromorphic architecture inspired directly by biological neural systems, using continuous-time dynamics, rather than the discrete, fixed-timestep…
lnn, neural architecture
Liquid Neural Networks (LNNs) are continuous-time recurrent networks with time-varying synaptic parameters inspired by C. elegans neural dynamics, enabling adaptive computation…
neural architecture
Liquid Time-Constant Networks (LTCs) are a class of continuous-time Recurrent Neural Networks (RNNs) — created by Ramin Hasani et al., where the hidden state's decay rate (time…
lis longest increasing subsequence, lis longest increasing, longest increasing subsequence, lis longest, longest increasing, increasing subsequence
LIS Longest Increasing Subsequence Introduction & Motivation LIS Longest Increasing Subsequence addresses a central problem in algorithms process control…
John McCarthy 1958, Recursive Functions of Symbolic Expressions 1960, cons cell, garbage collection, code as data, Scheme, Common Lisp, Lisp machine
Lisp is a programming language created by John McCarthy at MIT beginning in 1958, designed for symbolic computation and artificial intelligence research, in which programs…
linked list, car and cdr, symbolic processing, IPL, Lisp, list structure, FLPL
List processing is both a data structure and a programming paradigm: information is held in chains of cells, each linked to the next by a pointer, so that structures can grow…
audio & speech
Listen Attend Spell is a sequence-to-sequence speech-recognition model that maps audio features to text with attention - An encoder captures acoustic context, attention selects…
recommendation systems
ListNet is a listwise ranking method that optimizes probability distributions over ranked items. - It models ranking as distribution matching instead of independent pair…
Listwise Learning for Recommendations Introduction & Motivation Listwise Learning for Recommendations addresses a central problem in recsys systems and (target) optimization: how…
machine learning
Listwise ranking optimizes the entire ranked list — directly optimizing ranking metrics like NDCG or MAP rather than individual scores or pairs, the most sophisticated learning…
recommendation systems
Listwise Ranking is ranking optimization that models and optimizes the quality of full ranked lists - It aligns training more closely with user-facing recommendation outputs. What…
proxy, unified
LiteLLM is a Python library and proxy server that provides a unified OpenAI-compatible interface to 100+ LLM providers — enabling developers to switch between GPT-4, Claude…