mhtirogla
Home
Systems
API
Pricing
About
Log in
← Back
Home
›
Systems
›
AI and ML infrastructure
federated-learning
federated learning
Gboard
System
Techniques · 7
Primitives · 20
federated-learning
Dp sgd
ml-training · medium: Differentially private SGD: per-example gradient clipping plus Gaussian noise, with privacy accounting
Federated averaging
ml-training · medium: FedAvg: clients train locally, server averages weighted updates; handling non-IID data and stragglers
Differential privacy
security · medium: Calibrated noise (Laplace/Gaussian) and privacy budgets for aggregate releases
Privacy budget accounting
security · medium: Tracking cumulative epsilon/delta under composition (basic, advanced, Renyi/moments accountant)
Remote attestation
security · medium: Hardware-signed evidence (TPM quotes, SGX/TDX quotes, SEV-SNP reports, key attestation) proving device, key or enclave state to a verifier
Secure aggregation
security · medium: Summing client vectors so the server sees only the total: pairwise masks, secret-shared seeds, dropout recovery
Measured boot
security · medium: Each boot stage hashes the next into TPM PCRs (extend) and logs events for later attestation
SGD with momentum
ml-training · low: Gradient descent plus a decaying velocity buffer that averages minibatch noise and speeds up progress along shallow, consistent directions.
›
Gradient compression
ml-training · low: Reducing communicated updates by quantization, top-k sparsification and error feedback
ECDH key exchange (X25519, P-256)
security · low: Each side multiplies the other's public curve point by its own scalar and both get the same point; always validate the peer key and run the result through a KDF.
›
Shamir secret sharing (k-of-n)
security · low: Split a secret into n shares so any k reconstruct it and any k-1 reveal nothing at all, using a random polynomial over a finite field.
›
AdamW
ml-training · low: Adam's per-parameter adaptive steps with weight decay applied directly to the weights; the default optimizer for transformers and LLM pretraining.
›
AEAD (AES-GCM, ChaCha20-Poly1305)
security · low: Authenticated encryption that encrypts and detects tampering in one call; the only kind of symmetric encryption you should use.
›
Exponential backoff with jitter
systems · low: Retry after delays that grow geometrically and are randomized, so failed clients spread out instead of hammering a recovering service in synchronized waves.
›
Gradient clipping
ml-training · low: Clipping gradients by global norm (or value) to prevent loss spikes; interaction with mixed precision
Csprng
security · low: OS-backed cryptographically secure random numbers (getrandom), and when a fast PRNG is not enough
Weight synchronization
ml-training · low: Pushing updated policy weights from trainers to inference/actor workers (NCCL broadcast, sharded resharding)
Floating-point error
numerics · low: How IEEE 754 rounding error accumulates and how to contain it: compensated (Kahan/Neumaier) and pairwise summation, error-free transformations with FMA, ULP comparisons.
›
Conditional writes
distributed · low: Put-if-absent / if-match compare-and-swap on objects or keys as a commit primitive
Trusted execution environments
security · low: Enclaves/confidential VMs with remote attestation for computing on sealed data
X509 certificates
security · low: Certificate chains, path validation, SANs, revocation and short-lived issuance
Finite field arithmetic
numerics · low: Prime-field and extension-field arithmetic (Montgomery form, Goldilocks, BabyBear), vectorized
PRNGs (PCG, xoshiro)
numerics · low: Small, fast, statistically strong non-cryptographic generators: PCG32 and xoshiro256**, with seeding, parallel streams and unbiased ranges.
›
+ 4 more primitives
System
Techniques · 7
Primitives · 20
federated-learning
Dp sgd
ml-training · medium: Differentially private SGD: per-example gradient clipping plus Gaussian noise, with privacy accounting
Federated averaging
ml-training · medium: FedAvg: clients train locally, server averages weighted updates; handling non-IID data and stragglers
Differential privacy
security · medium: Calibrated noise (Laplace/Gaussian) and privacy budgets for aggregate releases
Privacy budget accounting
security · medium: Tracking cumulative epsilon/delta under composition (basic, advanced, Renyi/moments accountant)
Remote attestation
security · medium: Hardware-signed evidence (TPM quotes, SGX/TDX quotes, SEV-SNP reports, key attestation) proving device, key or enclave state to a verifier
Secure aggregation
security · medium: Summing client vectors so the server sees only the total: pairwise masks, secret-shared seeds, dropout recovery
Measured boot
security · medium: Each boot stage hashes the next into TPM PCRs (extend) and logs events for later attestation
SGD with momentum
ml-training · low: Gradient descent plus a decaying velocity buffer that averages minibatch noise and speeds up progress along shallow, consistent directions.
›
Gradient compression
ml-training · low: Reducing communicated updates by quantization, top-k sparsification and error feedback
ECDH key exchange (X25519, P-256)
security · low: Each side multiplies the other's public curve point by its own scalar and both get the same point; always validate the peer key and run the result through a KDF.
›
Shamir secret sharing (k-of-n)
security · low: Split a secret into n shares so any k reconstruct it and any k-1 reveal nothing at all, using a random polynomial over a finite field.
›
AdamW
ml-training · low: Adam's per-parameter adaptive steps with weight decay applied directly to the weights; the default optimizer for transformers and LLM pretraining.
›
AEAD (AES-GCM, ChaCha20-Poly1305)
security · low: Authenticated encryption that encrypts and detects tampering in one call; the only kind of symmetric encryption you should use.
›
Exponential backoff with jitter
systems · low: Retry after delays that grow geometrically and are randomized, so failed clients spread out instead of hammering a recovering service in synchronized waves.
›
Gradient clipping
ml-training · low: Clipping gradients by global norm (or value) to prevent loss spikes; interaction with mixed precision
Csprng
security · low: OS-backed cryptographically secure random numbers (getrandom), and when a fast PRNG is not enough
Weight synchronization
ml-training · low: Pushing updated policy weights from trainers to inference/actor workers (NCCL broadcast, sharded resharding)
Floating-point error
numerics · low: How IEEE 754 rounding error accumulates and how to contain it: compensated (Kahan/Neumaier) and pairwise summation, error-free transformations with FMA, ULP comparisons.
›
Conditional writes
distributed · low: Put-if-absent / if-match compare-and-swap on objects or keys as a commit primitive
Trusted execution environments
security · low: Enclaves/confidential VMs with remote attestation for computing on sealed data
X509 certificates
security · low: Certificate chains, path validation, SANs, revocation and short-lived issuance
Finite field arithmetic
numerics · low: Prime-field and extension-field arithmetic (Montgomery form, Goldilocks, BabyBear), vectorized
PRNGs (PCG, xoshiro)
numerics · low: Small, fast, statistically strong non-cryptographic generators: PCG32 and xoshiro256**, with seeding, parallel streams and unbiased ranges.
›
Cryptographic hashes (SHA-2, BLAKE3)
security · low: Collision-resistant fingerprints of data; SHA-256 for interoperability, BLAKE3 for speed, HMAC/HKDF for keyed uses.
›
Hardware backed keystore
security · low: Non-exportable keys in Secure Enclave, StrongBox, TPM: generation, use policies, biometric gating
Secure boot chain
security · low: Each boot stage verifies the signature of the next before running it, anchored in an immutable ROM key hash, with version counters against downgrade and a TPM log of what ran.
›
Constant time code
security · low: Writing secret-dependent code without branches or memory lookups on secrets; constant-time compare and select
Copy for AI
Similar systems
Share the most building blocks with this one.
differential-privacy
20 %