mirror of
https://github.com/harvard-edge/cs249r_book.git
synced 2026-11-09 10:39:41 -06:00
Branch commits skipped hooks; the full suite at the merge tip found these. - Bib: drop duplicate epochai2026benchmarks from the merge; rename villalobos2022will to villalobos2024will to match its year; re-record ledger hashes for 24 entries whose content is unchanged (tidy reformat) - Checks: scope cross-chapter-callout-ref to vol1 (vol2-4 carry 113 legacy refs); add Whetstone to heading proper nouns; lego-dead-code skips render output - Units: computed sizes in decimal units, hardware capacities through fmt_memory_capacity (ch6, ch11, ch14, ch16) - Prose: abbreviation first-use expansions after content moves, footnote marker placement, two footnote fixes, margin caption, bracketed source citations, index main, orphan and duplicate equation labels, dead LEGO exports - Auto-fixers: Black on hidden cells, whitespace, blank lines, tables, bibtex tidy, canonical spellings
794 lines
9.9 KiB
YAML
794 lines
9.9 KiB
YAML
# Heading & Callout Title Case Configuration for MIT Press Style (§10.3.1)
|
||
# Single source of truth for proper nouns, acronyms, compound terms, and exemptions
|
||
# across section headings (H3+) and callout titles (title="...").
|
||
|
||
acronyms_and_proper_nouns:
|
||
# General MLSys / AI & Engineering Domains
|
||
- AI
|
||
- ML
|
||
- AIOps
|
||
- MLOps
|
||
- DevOps
|
||
- ClinAIOps
|
||
- MLSys
|
||
- TinyML
|
||
- AutoML
|
||
- AutoAugment
|
||
- RandAugment
|
||
- NAS
|
||
|
||
# Processors, Accelerators & Memory Units
|
||
- GPU
|
||
- GPUs
|
||
- CPU
|
||
- CPUs
|
||
- TPU
|
||
- TPUs
|
||
- NPU
|
||
- DSP
|
||
- ASIC
|
||
- FPGA
|
||
- SoC
|
||
- MCU
|
||
- HBM
|
||
- HBM2
|
||
- HBM3
|
||
- HBM3e
|
||
- DRAM
|
||
- SRAM
|
||
- ROM
|
||
- EPROM
|
||
- EEPROM
|
||
- NVRAM
|
||
- VRAM
|
||
- NVMe
|
||
- PCIe
|
||
- PCI
|
||
- DDR4
|
||
- DDR5
|
||
- GDDR5
|
||
- GDDR6
|
||
- GDDR6X
|
||
- LPDDR4
|
||
- LPDDR5
|
||
- NAND
|
||
- NOR
|
||
- SLC
|
||
- MLC
|
||
- TLC
|
||
- QLC
|
||
|
||
# Interconnects & Networking
|
||
- NVLink
|
||
- NVSwitch
|
||
- GPUDirect
|
||
- DMA
|
||
- MMIO
|
||
- RDMA
|
||
- InfiniBand
|
||
- RoCE
|
||
- RoCEv2
|
||
- Clos
|
||
- Fat-Tree
|
||
- Dragonfly
|
||
- Torus
|
||
- Mesh
|
||
- Hypercube
|
||
- Butterfly
|
||
- Go-Back-N
|
||
|
||
# Compute Runtimes, Parallelism & Compilers
|
||
- CUDA
|
||
- ROCm
|
||
- OpenCL
|
||
- OpenMP
|
||
- MPI
|
||
- NCCL
|
||
- RCCL
|
||
- OpenACC
|
||
- BLAS
|
||
- cuBLAS
|
||
- cuDNN
|
||
- GEMM
|
||
- MAC
|
||
- MACs
|
||
- SIMD
|
||
- SIMT
|
||
- VLIW
|
||
- RISC
|
||
- CISC
|
||
- BatchNorm
|
||
- LayerNorm
|
||
- GroupNorm
|
||
- ReLU
|
||
- GELU
|
||
- SiLU
|
||
- PReLU
|
||
- LeakyReLU
|
||
- Sigmoid
|
||
- Softmax
|
||
- Tanh
|
||
- AllReduce
|
||
- AllGather
|
||
- ReduceScatter
|
||
- AllToAll
|
||
- Broadcast
|
||
- Scatter
|
||
- Gather
|
||
- FlashAttention
|
||
- PagedAttention
|
||
- RoPE
|
||
- ALiBi
|
||
- MoE
|
||
- KV
|
||
- DDP
|
||
- FSDP
|
||
- ZeRO
|
||
|
||
# Numerical Units, Precisions & Floating Point
|
||
- FLOP
|
||
- FLOPs
|
||
- FLOPS
|
||
- TFLOPs
|
||
- TFLOPS
|
||
- PFLOPs
|
||
- PFLOPS
|
||
- TOPS
|
||
- IOPS
|
||
- MIPS
|
||
- FP64
|
||
- FP32
|
||
- FP16
|
||
- FP8
|
||
- FP4
|
||
- BF16
|
||
- INT8
|
||
- INT4
|
||
- TF32
|
||
- NaN
|
||
- Inf
|
||
|
||
# Physical / Storage / Rates Units
|
||
- GB
|
||
- TB
|
||
- KB
|
||
- MB
|
||
- PB
|
||
- EB
|
||
- GiB
|
||
- TiB
|
||
- KiB
|
||
- MiB
|
||
- Gbps
|
||
- Mbps
|
||
- Tbps
|
||
- ns
|
||
- μs
|
||
- ms
|
||
- kHz
|
||
- MHz
|
||
- GHz
|
||
- THz
|
||
- Hz
|
||
- Wh
|
||
- kWh
|
||
- MWh
|
||
- W
|
||
- kW
|
||
- MW
|
||
|
||
# Systems, Protocols & Web/OS
|
||
- IO
|
||
- OS
|
||
- RTOS
|
||
- VM
|
||
- JVM
|
||
- JIT
|
||
- AOT
|
||
- RPC
|
||
- REST
|
||
- RESTful
|
||
- HTTP
|
||
- HTTPS
|
||
- TCP
|
||
- UDP
|
||
- IP
|
||
- DNS
|
||
- CDN
|
||
- TLS
|
||
- SSL
|
||
- SSH
|
||
- gRPC
|
||
- IPC
|
||
- UX
|
||
- UI
|
||
- API
|
||
- APIs
|
||
- SDK
|
||
- CLI
|
||
- IDE
|
||
- GUI
|
||
- SLA
|
||
- SLAs
|
||
- SLI
|
||
- SLO
|
||
- SLOs
|
||
- QPS
|
||
- RAID
|
||
- SSD
|
||
- HDD
|
||
- SRE
|
||
- QA
|
||
- KPI
|
||
|
||
# Legislation, Regulations & Standards
|
||
- HIPAA
|
||
- GDPR
|
||
- COPPA
|
||
- CCPA
|
||
- FERPA
|
||
- SOX
|
||
- HITRUST
|
||
- ISO
|
||
- SOC
|
||
- FIPS
|
||
- NIST
|
||
|
||
# Organizations, Conferences, Universities & Governments
|
||
- US
|
||
- UK
|
||
- EU
|
||
- USA
|
||
- FAA
|
||
- FDA
|
||
- IRB
|
||
- NHTSA
|
||
- DARPA
|
||
- NSF
|
||
- NIH
|
||
- IEEE
|
||
- ACM
|
||
- OSDI
|
||
- NSDI
|
||
- NeurIPS
|
||
- ICML
|
||
- ICLR
|
||
- CVPR
|
||
- ECCV
|
||
- ICCV
|
||
- ACL
|
||
- EMNLP
|
||
- NAACL
|
||
- KDD
|
||
- SIGMOD
|
||
- SIGCOMM
|
||
- SOSP
|
||
- FAST
|
||
- ATC
|
||
- HotOS
|
||
- MICRO
|
||
- ISCA
|
||
- HPCA
|
||
- SC
|
||
- MIT
|
||
- Stanford
|
||
- Berkeley
|
||
- CMU
|
||
- UIUC
|
||
- EPFL
|
||
- ETH
|
||
- Caltech
|
||
- Harvard
|
||
- Princeton
|
||
- Yale
|
||
- Columbia
|
||
|
||
# Companies & Hardware/Software Products
|
||
- NVIDIA
|
||
- AMD
|
||
- Intel
|
||
- ARM
|
||
- Apple
|
||
- Google
|
||
- Amazon
|
||
- Microsoft
|
||
- Meta
|
||
- Facebook
|
||
- IBM
|
||
- Qualcomm
|
||
- TSMC
|
||
- Samsung
|
||
- Uber
|
||
- Tesla
|
||
- Tempe
|
||
- DeepSeek
|
||
- Anthropic
|
||
- Cerebras
|
||
- Groq
|
||
- Graphcore
|
||
- SambaNova
|
||
- Tenstorrent
|
||
- Supermicro
|
||
- Oura
|
||
- Jeep
|
||
- Cherokee
|
||
- Rényi
|
||
- Jacobi
|
||
- Teton
|
||
- Delassus
|
||
- Clopper
|
||
- Pearson
|
||
- Euler
|
||
- Joule
|
||
- CLF
|
||
- CLFs
|
||
- CBF
|
||
- CBFs
|
||
- Slurm
|
||
# Research cluster schedulers (proper names)
|
||
- Tiresias
|
||
- Gandiva
|
||
- Themis
|
||
- Pollux
|
||
- PyTorch
|
||
- TensorFlow
|
||
- JAX
|
||
- Triton
|
||
- NumPy
|
||
- SciPy
|
||
- Pandas
|
||
- Linux
|
||
- POSIX
|
||
- FreeRTOS
|
||
- Zephyr
|
||
- XGBoost
|
||
- ONNX
|
||
- TensorRT
|
||
- TVM
|
||
- MLIR
|
||
- LLVM
|
||
- MLflow
|
||
- OpenAI
|
||
- DeepMind
|
||
- Horovod
|
||
- DeepSpeed
|
||
- Ulysses
|
||
- Megatron
|
||
- vLLM
|
||
- SGLang
|
||
- Michelangelo
|
||
- FBLearner
|
||
- Vertex
|
||
- Kubeflow
|
||
- Sagemaker
|
||
- MLPerf
|
||
- MLCommons
|
||
- LAPACK
|
||
- LINPACK
|
||
- Whetstone
|
||
- Roofline
|
||
- Instinct
|
||
- Xavier
|
||
- Orin
|
||
- Jetson
|
||
- Nsight
|
||
- Mirai
|
||
|
||
# Hardware Accelerator Models & Roman Numerals
|
||
- V100
|
||
- A100
|
||
- H100
|
||
- H200
|
||
- B100
|
||
- B200
|
||
- GB200
|
||
- T4
|
||
- L4
|
||
- L40
|
||
- RTX
|
||
- Xeon
|
||
- Epyc
|
||
- Ryzen
|
||
- M1
|
||
- M2
|
||
- M3
|
||
- M4
|
||
- Grace
|
||
- Hopper
|
||
- Blackwell
|
||
- Ampere
|
||
- Volta
|
||
- Pascal
|
||
- TPUv1
|
||
- TPUv2
|
||
- TPUv3
|
||
- TPUv4
|
||
- TPUv5
|
||
- TPUv5e
|
||
- TPUv5p
|
||
- I
|
||
- II
|
||
- III
|
||
- IV
|
||
- V
|
||
- VI
|
||
- VII
|
||
- VIII
|
||
- IX
|
||
- X
|
||
- XI
|
||
- XII
|
||
|
||
# Model Families & Algorithms
|
||
- CNN
|
||
- CNNs
|
||
- RNN
|
||
- RNNs
|
||
- LSTM
|
||
- GRU
|
||
- MLP
|
||
- MLPs
|
||
- NN
|
||
- DNN
|
||
- LLM
|
||
- LLMs
|
||
- ImageNet
|
||
- BERT
|
||
- GPT
|
||
- GPT-2
|
||
- GPT-3
|
||
- PaLM
|
||
- T5
|
||
- BART
|
||
- ViT
|
||
- CLIP
|
||
- ResNet
|
||
- AlexNet
|
||
- GoogLeNet
|
||
- MobileNet
|
||
- MobileNets
|
||
- EfficientNet
|
||
- VGG
|
||
- Inception
|
||
- YOLO
|
||
- DLRM
|
||
- BLOOM
|
||
- Gemini
|
||
- Claude
|
||
- Mistral
|
||
- Qwen
|
||
- Kaplan
|
||
- Chinchilla
|
||
- LLaMA
|
||
- Llama
|
||
- Adam
|
||
- AdamW
|
||
- SGD
|
||
- SGDM
|
||
- DP
|
||
- DP-SGD
|
||
- Adagrad
|
||
- RMSprop
|
||
- Lion
|
||
- Lamb
|
||
- Shampoo
|
||
- Muon
|
||
- DAgger
|
||
- DINOv2
|
||
- SAM
|
||
- ConceptGraphs
|
||
|
||
# Scholar Surnames & Named Theorems
|
||
- Margolus
|
||
- Levitin
|
||
- Landauer
|
||
- Wiener
|
||
- Moravec
|
||
- Brooks
|
||
- Pfeifer
|
||
- Bongard
|
||
- Ashby
|
||
- Hennessy
|
||
- Patterson
|
||
- Amdahl
|
||
- Gustafson
|
||
- Turing
|
||
- Sutton
|
||
- Karpathy
|
||
- Kuhn
|
||
- Goodhart
|
||
- Horowitz
|
||
- Williams
|
||
- Waterman
|
||
- Knuth
|
||
- Tanenbaum
|
||
- Dean
|
||
- Chintala
|
||
- Huang
|
||
- LeCun
|
||
- Han
|
||
- Reddi
|
||
- Stoica
|
||
- Huyen
|
||
- Emer
|
||
- Kullback
|
||
- Leibler
|
||
- Kolmogorov
|
||
- Smirnov
|
||
- KS
|
||
- Shannon
|
||
- Bayes
|
||
- Markov
|
||
- Bellman
|
||
- Boltzmann
|
||
- Poisson
|
||
- Taylor
|
||
- Young
|
||
- Daly
|
||
- Young-Daly
|
||
- Lagrange
|
||
- Doyle
|
||
- Kalman
|
||
- Lyapunov
|
||
# 2026-09-30 vol4 appendix headings: proper nouns the checker lowercased
|
||
- Riccati
|
||
- Hermite
|
||
- Euclidean
|
||
- Bernoulli
|
||
- Fisher
|
||
- Tippett
|
||
- Gnedenko
|
||
- "Bayes'"
|
||
|
||
- DAG
|
||
- ORM
|
||
- PRM
|
||
- Monte
|
||
- Carlo
|
||
|
||
# Agentic Systems: Runtimes, Sandboxes & Infrastructure (Vol III/IV)
|
||
- Docker
|
||
- Podman
|
||
- Kubernetes
|
||
- Firecracker
|
||
- Envoy
|
||
- Squid
|
||
- Ray
|
||
- Rust
|
||
- Devin
|
||
- SARATHI
|
||
- Stategraph
|
||
|
||
# Eponyms & Named Results (Vol III/IV)
|
||
- Markovian
|
||
- Pareto
|
||
- Hewitt-Agha
|
||
|
||
# Agentic & Physical AI: Additional Tools & Frameworks
|
||
- Tree-Sitter
|
||
- AST
|
||
- ASTs
|
||
- PageRank
|
||
- POSIX
|
||
- OAuth
|
||
- Kinesis
|
||
- DynamoDB
|
||
- Copilot
|
||
- EchoLeak
|
||
- SWE-bench
|
||
- WebArena
|
||
- Sagas
|
||
- MCP
|
||
- JSON-RPC
|
||
- DIFT
|
||
- W3C
|
||
- RLVR
|
||
- GRPO
|
||
- MCTS
|
||
- PRM
|
||
- PRMs
|
||
- CFG
|
||
- FSM
|
||
- ABI
|
||
- ABIs
|
||
- PCB
|
||
- PCBs
|
||
- ACB
|
||
- ACBs
|
||
- Bohrbug
|
||
- Bohrbugs
|
||
- Heisenbug
|
||
- Heisenbugs
|
||
- Protobuf
|
||
- Pickle
|
||
- Safetensors
|
||
- Tay
|
||
- Netflix
|
||
- Discord
|
||
- Zillow
|
||
- Gboard
|
||
- Epic
|
||
- Sepsis
|
||
- Knight
|
||
- Mars
|
||
- Orbiter
|
||
- Tesla
|
||
- Tempe
|
||
- Autopilot
|
||
- Williston
|
||
- Cruise
|
||
- Robotaxi
|
||
- VLA
|
||
- VLAs
|
||
- TSDF
|
||
- TSDFs
|
||
- WCET
|
||
- IMU
|
||
- IMUs
|
||
- PWM
|
||
- TCM
|
||
- ISO
|
||
- STO
|
||
- Simplex
|
||
- ODD
|
||
- ODDs
|
||
- MTBUF
|
||
- Toyota
|
||
- Ariane
|
||
- Pathfinder
|
||
- Challenger
|
||
- Shakey
|
||
# 2026-09-22: Vol IV war-story titles (ExoMars Schiaparelli, Deepwater Horizon) are proper nouns.
|
||
- Schiaparelli
|
||
- Deepwater
|
||
- Horizon
|
||
- Nyquist-Bode
|
||
- Nyquist–Bode
|
||
- Butler-Finelli
|
||
- Butler–Finelli
|
||
- LLaVA
|
||
- Qwen-VL
|
||
- RT-2
|
||
- OpenVLA
|
||
- Octo
|
||
- Jacobian
|
||
- Gaussian
|
||
- MicroVM
|
||
- Fermi
|
||
- BERT-Base
|
||
|
||
# Additional Eponyms & Named Results
|
||
- Heisenberg
|
||
- Dennard
|
||
- Jevons
|
||
- Hockney
|
||
- Baur
|
||
- Strassen
|
||
- Nyquist
|
||
- Bode
|
||
- Butler
|
||
- Finelli
|
||
- Bayesian
|
||
- Lie
|
||
- C²
|
||
- Postel
|
||
- Postel's
|
||
- Ousterhout
|
||
- Ousterhout's
|
||
- Birrell
|
||
- Nelson
|
||
- Saltzer
|
||
- Clark
|
||
- Denning
|
||
- Denning's
|
||
- Gray
|
||
- Gray's
|
||
- Sutton
|
||
- Sutton's
|
||
- Little
|
||
- Little's
|
||
- Mixtral
|
||
- Mixtral-8x7B
|
||
|
||
# Taxonomies & Custom Symbols
|
||
- D·A·M
|
||
- DAM
|
||
- C³
|
||
- C^3
|
||
- CCC
|
||
- C3
|
||
|
||
compound_names:
|
||
- [Monte, Carlo]
|
||
- [Two, Generals]
|
||
- [Claude, Code]
|
||
- [MLPerf, Inference]
|
||
- [Fleet, Law]
|
||
- [MLPerf, Training]
|
||
- [MLPerf, Tiny]
|
||
- [MLPerf, Mobile]
|
||
- [MLPerf, Client]
|
||
- [MLPerf, HPC]
|
||
- [Oura, Ring]
|
||
- [Apple, Watch]
|
||
- [Apple, Silicon]
|
||
- [Google, Glass]
|
||
- [Tensor, Core]
|
||
- [Tensor, Cores]
|
||
- [EU, Act]
|
||
- [AI, Act]
|
||
- [Grand, Teton]
|
||
- [Jeep, Cherokee]
|
||
- [Meta, Grand]
|
||
- [FBLearner, Flow]
|
||
- [Vertex, AI]
|
||
- [Amazon, SageMaker]
|
||
- [Nsight, Systems]
|
||
- [Nsight, Compute]
|
||
|
||
- [Model, Context]
|
||
- [Context, Protocol]
|
||
- [Agent, Control]
|
||
- [Control, Block]
|
||
- [Knight, Capital]
|
||
- [Zillow, Offers]
|
||
- [Epic, Sepsis]
|
||
- [Mars, Climate]
|
||
- [Climate, Orbiter]
|
||
- [Tesla, Autopilot]
|
||
- [Cruise, Robotaxi]
|
||
- [News, Feed]
|
||
- [Jevons, Paradox]
|
||
- [GPUDirect, Storage]
|
||
- [Discord, Read]
|
||
- [Read, States]
|
||
- [Federated, MobileNet]
|
||
- [at, Scale]
|
||
- [Apple, Card]
|
||
|
||
dam_axes:
|
||
- Data
|
||
- Algorithm
|
||
- Machine
|
||
- Compute
|
||
- Computation
|
||
- Communication
|
||
- Coordination
|
||
|
||
concept_terms_lower:
|
||
- iron law
|
||
- degradation equation
|
||
- verification gap
|
||
- bitter lesson
|
||
- ml node
|
||
- data wall
|
||
- compute wall
|
||
- memory wall
|
||
- power wall
|
||
- energy corollary
|
||
- machine learning operations
|
||
- transformer
|
||
- four pillars framework
|
||
- scaling laws
|
||
- information roofline
|
||
- long tail
|
||
- data gravity
|
||
- napkin math
|
||
- starving accelerator
|
||
- latency cliff
|
||
- roofline model
|
||
|
||
# ─── Exemptions ─────────────────────────────────────────────────────────────
|
||
# Full headings exempt from case transformation:
|
||
skip_headings:
|
||
- "Duplex transport framing: stdio vs. SSE"
|
||
- "Idempotency keys and the Two Generals' problem"
|
||
- "1-bit Adam: Compression-aware optimization"
|
||
- "Empirical validation: 50-layer comparison"
|
||
|
||
# Full callout titles exempt from case transformation:
|
||
skip_callout_titles: []
|
||
|
||
# Full titles exempt across BOTH section headings and callout titles:
|
||
skip_titles: []
|