ARIS-in-AI-Offer (ARIS in 秋招)
Hoping to make your 秋招 (qiūzhāo, Chinese AI campus recruiting season) a little easier 🌱
📖 中文版 (Chinese version): README_CN.md
📱 One link for everything: easyaioffer.github.io (short address; the site itself is wanshuiyin.github.io/ARIS-in-AI-Offer) — every sheet on one page, searchable, 中/EN switch, dark mode, and a per-reader 已读 tracker. Save it on your phone.
📚 Jump to a topic — 35 first-party cheat sheets across 8 categories, plus a community-contributed question bank:
🧠 General / Foundations · 🎯 Post-Training & Reasoning · 🏛️ LLM Architecture & Systems · 🌊 Generative Models — Theory & Tokenizers · 🎨 Generation Systems (Image / Video / 3D / Diffusion Post-Training) · 👁️ Multimodal · 🤖 Agents · 🦾 Embodied AI & World Models / 具身与世界模型
Or browse the full 📚 Tutorial Index ↓ · jump to 🌐 ARIS-Homepage ↓.
· 
· 
· 
· 
· 
🏆 Built on a battle-tested foundation — the ARIS main repo has ~10k GitHub stars, was HuggingFace Daily Papers #1, won AI Digital Crew Project of the Day, and ships 74+ research skills across 7+ platforms. This isn’t a vaporware preview — every cheat sheet here is the production output of the same /interview-cheatsheet + /render-html workflow used in academic-research production.
A curated, bilingual (中文 + English) collection of ML / LLM / multimodal / diffusion / agent / generative-model interview cheat sheets, auto-generated by the ARIS — Auto Research in Sleep /render-html workflow.
Each cheat sheet is a long-form Chinese tutorial with: formula derivations · from-scratch PyTorch code · 25 high-frequency interview questions (L1 essentials · L2 advanced · L3 top-tier lab).
📖 Preview (above): one snapshot per pillar, taken from the Diffusion Foundations cheat sheet — ① Foundations (formula derivations + intuition + TL;DR), ② Interview Q&A (25 high-frequency questions stratified L1/L2/L3), ③ From-Scratch Code (runnable PyTorch, including CFG training + DDIM sampling). Every cheat sheet in this collection follows the same three-pillar structure.
🌐 ARIS-Homepage preview — CV → fact-checked academic homepage (click to expand)
Same /render-html workflow turning a CV into a fact-checked academic homepage. Live demo at wanshuiyin.github.io. Details + pipeline diagram in the ARIS-Homepage section ↓.
📝 Long-form blog preview — A Survey on Continuous DLM (2026 H1) (click to expand)
A standalone hand-authored long-form technical survey. A Survey on Continuous DLM (2026 H1, 6 papers) — Chinese-language survey by Ruofeng Yang (SJTU), written end-to-end via cross-model discussion (Claude Opus 4.7 + Codex GPT-5.5 xhigh + Gemini auto-gemini-3). 📖 Read full blog ↗.
📱 HTML reads cleanly everywhere
Phone on the subway, iPad at a café, laptop in the library — same HTML link opens equally well:
- 🧮 MathJax renders all LaTeX formulas (not screenshots — scalable, copyable, selectable)
- 💻 highlight.js colors all PyTorch code blocks
- 📐 Responsive layout adapts to any window width — no overflow, no blur
- 📑 Sticky TOC for jumping around long documents
- 💾 Single-file HTML — download once, read offline, no backend required
📢 What’s New
- 2026-09-11 —
🌍 #35 World Models + a 2025–26 refresh of 3D Generation and Quantization — from community feedback (“3D 和 WM 有点少/老”, “quant 主流用 PTQ 和 QAT”). World Models is new and deliberately curated — influential, representative, major-lab, open-weight work only, with the selection rule stated in the sheet and the famous-but-excluded listed with reasons: latent dynamics (RSSM → Dreamer v1–4, MuZero, TD-MPC2) · JEPA (I-/V-JEPA, V-JEPA 2/2.1, DINO-WM) · generative/interactive (Genie → Genie 3, GameNGen, Cosmos 1→3, Matrix-Game 3.0, Atlas) · three embodied routes · evaluation and the definition fight. 3D gains feed-forward reconstruction (DUSt3R → VGGT → MapAnything / DA3, with the confidence-loss derivation), feed-forward 3DGS, native mesh generation (MeshGPT → Meshtron, tokenization budgets), TRELLIS.2 / Hunyuan3D 2.x on a latent-representation axis, part-level and sim-ready assets; five duplicated L3 questions replaced. Quantization §10 rewritten around QAT: PACT/LSQ derivations, ParetoQ, Gemma 3/4 official QAT weights, Apple’s 2-bpw on-device stack, DeepSeek FP8 / NVFP4 / MXFP8 training recipes — and the 2026 PTQ-vs-QAT guidance with costs cited rather than asserted. All bilingual; new runnable code/world_models_toy.py (RSSM balanced-KL fixed point, JEPA shortcut, gridworld LAM — asserts against analytic answers). Research fanned out to Opus 5 with every arXiv ID fetched; drafts reviewed by gpt-6-astra xhigh in 2–3 rounds each. world_models_tutorial.html.
- 2026-09-10 — 🎯 #34 Modern Diffusion Post-Training: Flow-GRPO · DGPO · DiffusionNFT — the three 2025 online-RL methods for flow-matching T2I, organised around the one thing that separates them: how the reward enters the gradient — per-step policy gradient on a marginal-preserving SDE, group-level DPO with the $\log Z$ cancellation, or a reflected two-branch regression on the forward process. Flow-GRPO moves here out of the Diffusion Post-Training sheet, with its reverse-SDE sign and closed-form KL corrected on the way. Bilingual, with a runnable script (
code/diffusion_online_rl.py — analytic checks of marginal preservation, the NFT update sign, and DGPO’s balanced weights) and 25 高频题. Design-reviewed before drafting (64 guardrails) plus three review rounds, all gpt-6-astra xhigh. modern_diffusion_post_training_tutorial.html.
📋 Earlier updates (22)
- 2026-07-31 —
✂️ Interview-scannability overhaul for #30-33 + collection-wide table-scroll fix — responding to reader feedback (“comprehensive but unscannable before an interview”): Transformer Block restructured — the concrete block assembly + dataflow + runnable code moved from a buried §7 up to §1 (answer first, rationale after); all four de-hedged — lead sentence states the conclusion, qualifiers demoted to notes, one canonical home per caveat (inference visible prose -33.5%, transformer_block -17.2%, eval -17.4%), zero factual changes; a quantitative screen confirmed the older 29 tutorials healthy. Renderer wide-table overflow fixed, all 67 HTMLs re-rendered; all four EN editions retranslated + fidelity-reviewed. Gate PASS. (#37 · 28e4fb9)
- 2026-07-22 — 🧱 4 new cheat sheets (#30-33): Transformer Block · LLM Evaluation & Benchmarking · LLM Pretraining Pipeline · LLM Inference & Serving Stack — four foundational/systems topics in one batch: residual topologies & the MQA/GQA/MLA design axes, unbiased pass@k estimation & LLM-as-judge, Kaplan vs Chinchilla & the corpus-factory data pipeline, the request state machine & PagedAttention/KV lifecycle — 25-30 interview questions each. First batch to move the cross-model design review before drafting (90-105 numbered guardrails each), then 3-5 independent GPT-5.6-sol review batches; 3 real bugs were each caught independently by multiple batches (Bradley-Terry separation criterion, the best-of-n KV formula). All bilingual with runnable scripts, gate PASS. (#36 · 74644cf)
- 2026-07-13 — 🔍 Diffusion-cluster resweep completes the sweep — 8 tutorials, 66 fixes, all 28 tutorials now under GPT-5.6-sol — covers the diffusion/generative-media cluster deferred from the prior resweep (07-12). Same two-stage pipeline (GPT-5.6-sol finds → independent Claude adversarially verifies): 67 candidates → 66 fixed, 2 REFUTED left untouched; includes 5 real code bugs (iCT/FSQ/LFQ/DDPO) and a recurring “conditional path is straight” ≠ “marginal ODE trajectory is straight” confusion. Gate PASS. (#32 · bfae8f1)
- 2026-07-12 — 🔍 Full-collection resweep — 20 tutorials, 236 fixes, cross-model review upgraded to GPT-5.6-sol — after the reviewer moved from GPT-5.5 to GPT-5.6-sol, re-audited training fundamentals / attention / RLHF / inference systems / PEFT / agents / RAG·VLM (20 files). Same two-stage pipeline: 240 candidates → 236 fixed, 4 REFUTED left untouched; includes 5 real code bugs and StarPO’s acronym settled from the paper’s own abstract. Gate PASS. (#31 · 59636aa)
- 2026-07-11 — 🔤 #29 Tokenization cheat sheet — fills the foundation the collection had been taking for granted: BPE / WordPiece / Unigram (training ≠ encoding trap) · byte-level BPE vs byte fallback · the easily-confused vocab quantities · the tokenizer-free frontier. Bilingual, with a runnable script (
code/tokenization.py — exact GPT-2 bytes_to_unicode, Unigram DPs hand-checked, pure stdlib, verified on a real box) and 25 高频题. First tutorial with cross-model review moved up to the design stage (81 guardrails before drafting + 5 review batches + an ultra final pass, corrections flowing both ways). tokenization_tutorial.html.
- 2026-06-29 — ⚙️ #28 Optimizers & LR Schedules cheat sheet — the optimizer/schedule training-mechanics gap: SGD·Momentum·Nesterov · AdaGrad→RMSProp→Adam (the adaptive lineage) · Adam bias correction (the uncorrected first step is ~3.16× too large, not small — the direction trap) · AdamW decoupled weight decay (Adam’s L2 ≠ weight decay) · frontier (Muon / Lion / Shampoo / SOAP / Adafactor / LAMB / Sophia) · LR schedules (warmup / cosine / Noam / WSD / one-cycle) · weight decay + LR-batch scaling + no-decay groups · grad clipping + LLM hyperparams (β₂=0.95). Bilingual (中文 + EN), with a runnable script (
code/optimizer_lr_schedule.py — SGD/Adam/AdamW from-scratch vs torch.optim, the executable AdamW≠Adam+L2 proof, the bias-correction direction, cosine-warmup, verified on a real box) and 25 高频题. optimizer_lr_schedule_tutorial.html.
- 2026-06-19 — 🧱 #27 Normalization / Residual / Init cheat sheet — the foundational training-mechanics hole: BatchNorm / LayerNorm / RMSNorm / GroupNorm · Pre-vs-Post-LN (Xiong’s gradient argument — why Post-LN needs warmup) · DeepNorm / Sandwich / QK-Norm · residual connections + scaling (LayerScale / ReZero / GPT-2’s 1/√(2N)) · Xavier vs Kaiming (the ReLU factor-2, and why the preserved quantity is the second moment E[y²], not Var) · μP (width-invariant HP transfer) · Fixup / NFNets / DyT (norm-free) · plus the covariate-shift debunk (Santurkar). Bilingual (中文 + EN), with a runnable script (
code/normalization.py — from-scratch LN/RMSNorm vs torch + Pre/Post-LN gradient + Kaiming second-moment + GPT-2 residual scaling, verified on a real box) and 25 高频题. The cross-model review caught a real methodological confound in the Pre/Post-LN gradient demo (an output-normalization artifact) and redesigned it to a loss-robust top/bottom weight-grad ratio. normalization_init_tutorial.html.
- 2026-06-18 — ⚡ #26 Linear / Sparse Attention cheat sheet — the sub-quadratic / efficient-attention hole the collection kept referencing but never derived: linear attention (kernel φ + associativity → matrix-state RNN) · SSM / Mamba (selective S6) · Mamba-2 / SSD (1-semiseparable duality ≡ structured masked linear attention) · DeltaNet / Gated DeltaNet (overwrite update) · chunkwise-parallel training · trainable sparse (NSA three-branch / MoBA / Lightning / DSA) · hybrid (Jamba / Hymba / Qwen3-Next / Kimi-Linear / MiniMax-01). Bilingual (中文 + EN), with a runnable script (
code/linear_sparse_attention.py — chunkwise ≡ recurrent equivalence + delta-rule + block-sparse, verified on a real box) and 25 高频题, settled through multi-batch Codex GPT-5.5 xhigh citation / math / code / answer / overall + render-fidelity review. linear_sparse_attention_tutorial.html (#22 · b84e913).
- 2026-06-13 → 06-14 — 🧩 Two must-know cheat sheets shipped: #24 LoRA / PEFT + #25 RAG + Embedding / Retrieval — the two glaring holes the collection used everywhere but never derived. LoRA/PEFT: B=0 identity start · α/r vs rsLoRA √r · zero-latency merge · QLoRA (NF4 / double-quant / paged) · DoRA · the family vs Adapter / Prefix / Prompt / BitFit. RAG: bi-encoder · InfoNCE + hard negatives · Matryoshka · BM25 · HNSW · RRF hybrid · cross-encoder vs ColBERT late interaction · HyDE / Self-RAG / CRAG · GraphRAG · RAGAS. Both bilingual (中文 + EN), each with from-scratch PyTorch, a runnable script (
code/lora.py · code/rag_embedding.py, verified on a real box) and 25 高频题, settled through multi-round Codex GPT-5.5 xhigh math/code + render + EN translation-fidelity review. lora_peft_tutorial.html · rag_embedding_retrieval_tutorial.html.
- 2026-06-08 —
🔧 tools/render_html.py now strips a leading UTF-8 BOM before frontmatter detection (6cc4876).
- 2026-06-05 —
🔭 Community Showcase: an online tutorial-collection site by @QiZishi — merges all 23 Chinese tutorials here with Datawhale Hello-Agents’ LLM interview Q&A into one card-index reading site (read online · from #3). The README gains a Community Showcase section listing community derivatives — build something on top of these tutorials and open an issue to get it listed (f8e7d33).
- 2026-06-02 — 📝 Two new blogs — Continuous DLM (representation perspective, v2) + Diffusion × Representation × Manifold — the Continuous DLM survey is upgraded to a representation-perspective expansion (replacing the earlier v1), and a companion blog traces the “borrow representation / use the manifold” threads across image & video diffusion (SSL / Consistency / REPA / RAE / JiT / V-JEPA2). Both hand-authored, cross-model reviewed; the two cross-reference each other. Live:
continuous_dlm_representation_perspective.html · diffusion_representation_manifold.html. See the Blog Index ↓.
- 2026-06-01 — 📝 New blog: a deep-dive guide to NVIDIA Cosmos 3 (omnimodal world model · MoT architecture) — a long-form Chinese popsci walkthrough of the 138-page Cosmos 3 technical report (15 sections · 12 figures): how understanding and generation are stitched into one Transformer (Mixture-of-Transformers), the training recipe, scaling / serving, and the three model sizes. Reviewed end-to-end across 5 rounds of Codex GPT-5.5 xhigh cross-model audit (numbers + mechanisms checked, overclaims trimmed). Self-contained single-file HTML at
docs/blogs/cosmos3_mot_guide.html — a hand-authored guide (ELF blog format), not part of the audited /render-html pipeline. All figures and numbers are from the NVIDIA Cosmos 3 report (github.com/nvidia/cosmos); © the original authors, used here as an attributed popsci guide.
- 2026-05-31 — 🔍 Cross-model audit pass — real errors caught & fixed across the whole collection, now CI-enforced — a fresh Codex GPT-5.5 xhigh re-review of every cheat sheet (中文 + EN) surfaced genuine technical mistakes the first pass had missed, and fixed them: DeepSeek-V3’s FP8 GEMM accumulates in FP32, not bf16; Qwen2-VL M-RoPE base is 1e6; Molmo’s vision tower is CLIP, not SigLIP; TensoRF complexity is O(N³); a broken StreamingLLM render callout; plus 10 EN translation-fidelity fixes (meaning reversals, a dropped framework list). Every
docs/ artifact now carries a traceable cross-model review, and a new CI gate (tools/verify_reviews.py --mode strict --reproduce, wired in review-audit.yml) blocks any PR whose HTML isn’t a reviewed view that byte-reproduces from its source — so no un-reviewed or hand-edited content can land. Also merged today: a Focus-dim reading mode for --blog-mode pages (dim non-current sections · floating 🎯 toggle · ↑/↓ section nav), reflowed from the dllm blog effort into the academic template — dormant for the cheat sheets (gated under .aris-blog), it activates in blog-mode, ready for the first blogs rendered through it (landing soon). (5aae952 · 498ecf2)
- 2026-05-28 — 📝 First blog shipped: A Survey on Continuous DLM (2026 H1, 6 papers) — long-form Chinese technical survey by Ruofeng Yang (SJTU); superseded by the v2 rewrite at the same path, see the 2026-06-02 entry, written end-to-end through cross-model discussion (Claude Opus 4.7 + Codex GPT-5.5 xhigh + Gemini auto-gemini-3). Compares ELF, ByteDance Cola-DLM, and Flow-Matching family across discrete-DLM problems, the “known-unknown” continuous space idea, training pipeline, architecture / params / shapes, inference grids + Tab 6/7 numerical results, denoising trajectories, and a Field Landscape against Cola-DLM. Lives at
docs/blogs/continuous_dlm_representation_perspective.html (1.7 MB self-contained, no build) — a hand-authored long-form HTML, outside the audited /render-html pipeline. Preview strip + live link at the top of this README. (8475a2d)
- 2026-05-28 — ⚡
render-html P0 polish + all 23 tutorials regenerated — academic template gained 7 interactive features: print degradation fix (PDF no longer loses <details> content), TOC sidebar scrollspy (current section auto-highlights as you scroll), figure lightbox (native <dialog> with focus trap + Esc), long-code auto-collapse (<pre> ≥30 lines wrapped in <details class="code-card">, per-block override via ```python {collapsed} / {open} fence flags), paper citation popover (new [[key]] MD syntax + --papers <papers.json> sidecar), eyebrow cleanup (marketing uppercase → body-serif gray), --blog-mode infrastructure (opt-in aris-blog body class). XSS-hardened script injection via json_for_script() (escapes </script> break-out). All 23 bilingual tutorial pairs (= 46 HTMLs) regenerated to pick up the new template shell — source MDs untouched. Codex GPT-5.5 xhigh 4-round review (design × 2 → code × 1 → spot-check × 1). Try it: scroll attention_tutorial.html and watch the TOC sidebar follow (b79c57d, 8793f40).
- 2026-05-26 — 🐍 5 runnable PyTorch tutorial scripts — first runnable-code contribution in
docs/tutorials/code/: mha.py (MHA + causal mask) · axial_attention.py (H/W axial + complexity table) · flow_matching.py (Rectified Flow on 2D moons) · mmdit_block.py (double-stream MMDiT block) · toy_mmdit_t2i_pipeline.py (end-to-end T2I skeleton). Pure PyTorch, CPU-runnable in seconds, every script ships with built-in assert sanity checks (shape parity, numerical agreement with nn.MultiheadAttention where applicable). Pairs with attention_tutorial.md / flow_matching_tutorial.md / image_generation_systems_tutorial.md (f63f468).
- 2026-05-24 — 🐙 ARIS-Homepage v1.1:
--from-repos — snapshot user-selected owner/repo list via gh CLI; LLM agent merges repo timelines into homepage News + featured_projects[].github. Private repos skipped by default. Closes #2 by @Yafei-Liu99 (cdcf9a2).
- 2026-05-23 — 🌐 ARIS-Homepage v1 shipped — CV → fact-checked academic homepage (DBLP audit blocks wrong venue / year). Single-file HTML; Codex / Gemini reviews optional. Live demo: wanshuiyin.github.io. Skill:
skills/homepage-generator/SKILL.md (b818c1d).
- 2026-05-22 — 🦾 Featured community contribution: 具身智能高频面试题库 by @WinstonJQ — 413 questions across 8 卷 (VLA / 模仿学习 / RL / 世界模型 / 工程落地 / 腿足控制 / 3D 感知 / 系统设计). Hosted externally; linked from the new “🦾 Embodied AI” category in the Tutorial Index (b1ebb6f).
- 2026-05 — 📚 4 new bilingual cheat sheets: KL Divergence in RLHF (k1/k2/k3 · placement gradient bias), LLM On-Policy Distillation (MiniLLM / GKD / Qwen3 / Tinker), Diffusion Post-Training (DDPO / DPOK / DRaFT / AlignProp / Diffusion-DPO / Flow-GRPO), Diffusion / Flow Distillation (CM / iCT / sCM / CTM / LCM / DMD/DMD2 / ADD/LADD). Total now: 23 first-party cheat sheets.
- 2026-05 —
📖 README restructure — preview-strip banner, ARIS credentials at top (badges + 10K-star foundation paragraph), shared WeChat community QR with the main ARIS repo.
📝 Blog Index
Long-form technical blogs — hand-authored, cross-model reviewed; outside the audited /render-html pipeline (figures © their original authors, used with attribution).
| Blog |
What it covers |
|
| NVIDIA Cosmos 3 — MoT Architecture Deep-Dive (中文) |
Omnimodal world model · Mixture-of-Transformers · a walkthrough of the 138-page Cosmos 3 technical report |
📄 Read |
| A Survey on Continuous DLM — Representation Perspective (中文) |
Continuous diffusion language models through a representation lens · ELF / ByteDance Cola-DLM / Flow-Matching family (2026 H1) |
📄 Read |
| Diffusion × Representation × Manifold (中文) |
The “borrow representation / use the manifold” threads in image & video diffusion · SSL / Consistency / REPA / RAE / JiT / V-JEPA2 · cross-referenced with the Continuous DLM survey |
📄 Read |
📚 Tutorial Index
🌐 Bilingual editions: every cheat sheet ships with both a Chinese (default) and an English HTML — filenames are *_tutorial.html (CN) and *_tutorial_en.html (EN). HTML columns below link to both.
🧠 General / Foundations
| Topic |
HTML 中文 |
HTML EN |
MD |
| Attention Interview Cheat Sheet |
📄 CN |
📄 EN |
MD |
| Transformer Block (Post-LN/Pre-LN/branch pre+post residual topologies · MHA/MQA/GQA/MLA · Dense FFN vs MoE · GPT-2→Llama-style evolution) |
📄 CN |
📄 EN |
MD |
| LLM Evaluation & Benchmarking (pass@k unbiased estimation · evaluator ladder · benchmark contamination detection · LLM-as-judge · Bradley-Terry/Elo) |
📄 CN |
📄 EN |
MD |
| Normalization / Residual / Init (BatchNorm / LayerNorm / RMSNorm / Pre-vs-Post-LN / DeepNorm / QK-Norm / Xavier·Kaiming / μP) |
📄 CN |
📄 EN |
MD |
| Optimizers & LR Schedules (SGD·Momentum / Adam·AdamW / Muon·Lion·Shampoo·SOAP / warmup·cosine·WSD) |
📄 CN |
📄 EN |
MD |
| Tokenization (BPE / WordPiece / Unigram·SentencePiece / byte-level·byte fallback / vocab·fertility·BPB / tokenizer-free) |
📄 CN |
📄 EN |
MD |
| KL Divergence in RLHF (k1/k2/k3 · placement gradient bias) |
📄 CN |
📄 EN |
MD |
🎯 Post-Training & Reasoning
| Topic |
HTML 中文 |
HTML EN |
MD |
| RLHF / DPO / GRPO / PPO |
📄 CN |
📄 EN |
MD |
| Reasoning Models (o1 / R1 / Test-Time Compute / PRM) |
📄 CN |
📄 EN |
MD |
| LLM On-Policy Distillation (MiniLLM / GKD / Qwen3 / Tinker) |
📄 CN |
📄 EN |
MD |
| LoRA / PEFT (LoRA / QLoRA / DoRA / rsLoRA / PiSSA / AdaLoRA / (IA)³) |
📄 CN |
📄 EN |
MD |
🏛️ LLM Architecture & Systems
| Topic |
HTML 中文 |
HTML EN |
MD |
| MoE (DeepSeek-V3 / Mixtral / Llama 4) |
📄 CN |
📄 EN |
MD |
| Long Context (RoPE / YaRN / NTK / MLA / StreamingLLM) |
📄 CN |
📄 EN |
MD |
| Linear / Sparse Attention (Linear Attn / SSM·Mamba / Mamba-2·SSD / DeltaNet / NSA·MoBA / Hybrid) |
📄 CN |
📄 EN |
MD |
| KV Cache + Speculative Decoding (Medusa / EAGLE / MLA) |
📄 CN |
📄 EN |
MD |
| Quantization (GPTQ / AWQ / SmoothQuant / FlatQuant · FP8 / NVFP4 · QAT: LSQ / ParetoQ / Gemma-QAT · FP4 training) |
📄 CN |
📄 EN |
MD |
| Distributed Training (DDP / FSDP2 / ZeRO / TP / PP / EP / SP) |
📄 CN |
📄 EN |
MD |
| LLM Pretraining Pipeline (Kaplan vs Chinchilla scaling laws · corpus-factory data pipeline · document packing/loss masking · checkpoint resume) |
📄 CN |
📄 EN |
MD |
| LLM Inference & Serving Stack (request state machine · exact sampling-operator definitions · PagedAttention/KV lifecycle · continuous batching/chunked prefill · disaggregation) |
📄 CN |
📄 EN |
MD |
🌊 Generative Models — Theory & Tokenizers
| Topic |
HTML 中文 |
HTML EN |
MD |
| Flow Matching Quick Reference |
📄 CN |
📄 EN |
MD |
| Diffusion Foundations (DDPM / Score / DDIM / EDM / CFG) |
📄 CN |
📄 EN |
MD |
| VAE / VQ-VAE / VQ-GAN / FSQ |
📄 CN |
📄 EN |
MD |
🎨 Generation Systems — Image / Video / 3D / Diffusion Post-Training
| Topic |
HTML 中文 |
HTML EN |
MD |
| Image Gen Systems (LDM / SD / SDXL / SD3 / FLUX / ControlNet) |
📄 CN |
📄 EN |
MD |
| Video Gen (Sora / Hunyuan-Video / Kling / Wan / Movie Gen) |
📄 CN |
📄 EN |
MD |
| 3D Gen (NeRF / Instant-NGP / 3DGS / SDS · DUSt3R→VGGT feed-forward · native mesh · TRELLIS / Hunyuan3D 2.x · part-level / sim-ready) |
📄 CN |
📄 EN |
MD |
| Diffusion Post-Training (DDPO / DPOK / DRaFT / AlignProp / Diffusion-DPO) |
📄 CN |
📄 EN |
MD |
| Modern Diffusion Post-Training (Flow-GRPO / DGPO / DiffusionNFT) |
📄 CN |
📄 EN |
MD |
| Diffusion / Flow Distillation (CM / iCT / sCM / CTM / LCM / DMD/DMD2 / ADD/LADD) |
📄 CN |
📄 EN |
MD |
👁️ Multimodal
| Topic |
HTML 中文 |
HTML EN |
MD |
| VLM (CLIP / LLaVA / Qwen-VL / DeepSeek-VL) |
📄 CN |
📄 EN |
MD |
🤖 Agents
| Topic |
HTML 中文 |
HTML EN |
MD |
| Agent Foundations (ReAct / MCP / A2A / SWE-bench / GAIA / OSWorld) |
📄 CN |
📄 EN |
MD |
| Agentic RL (AgentTuning / ToolRL / RAGEN / WebRL / SWE-RL / GRPO for tool use) |
📄 CN |
📄 EN |
MD |
| Multi-Agent & Long-Horizon (CAMEL / AutoGen / MetaGPT / MoA / Debate / MemGPT / LATS) |
📄 CN |
📄 EN |
MD |
| Self-Evolving Agents (Ctx2Skill / Native Evolution / A²RD / Voyager / Reflexion / STaR) |
📄 CN |
📄 EN |
MD |
| RAG + Embedding / Retrieval (InfoNCE / 难负例 / Matryoshka / BM25 / RRF / ColBERT / GraphRAG) |
📄 CN |
📄 EN |
MD |
🎉 23 tutorials live (bilingual) (2026-05) — each ships with both Chinese and English HTML. Seven buckets: General · Post-Training · Architecture · Generative · Multimodal · Agents · Diffusion Post-Training. This round adds 4 new sheets: KL Divergence in RLHF, LLM On-Policy Distillation, Diffusion Post-Training, Diffusion Distillation. More (Flow-OPD / Audio Gen / further SOTA updates) coming — PRs welcome (see CONTRIBUTING).
🦾 Embodied AI & World Models / 具身与世界模型
| Topic |
HTML 中文 |
HTML EN |
MD |
| World Models (Dreamer / MuZero / TD-MPC2 · JEPA / V-JEPA 2 · Genie / GameNGen / Cosmos 3 / Matrix-Game / Atlas · embodied use · evaluation) |
📄 CN |
📄 EN |
MD |
🌟 Community contribution by @WinstonJQ — hosted externally on a separate repo, generously shared with the community. If it helps your interview prep, please ⭐ the source repo to thank the author 🙏
🤖 How These Are Generated
Every tutorial uses ARIS’s /interview-cheatsheet skill:
- Plan — 12-14 sections (TL;DR · Intuition · Formulas · Code · Variants · Complexity · 25 Q&A)
- Draft — 600-1000 lines of Chinese tutorial + runnable from-scratch PyTorch
- Cross-model review — fresh-thread codex GPT-5.5 xhigh audit on 10 properties (formula correctness · code runnability · citation accuracy · table-pipe escapes · callout style · personal-info leak · …)
- Fix loop — trajectory-based; keep going if FAIL set is shrinking, stop if same issue recurs or ~6 rounds without convergence
/render-html — single-file HTML render + 13-property render audit (information fidelity · TOC · math · code highlight · safety · privacy · …).review.json — full audit trail saved next to each tutorial
Cross-model adversarial review (executor ≠ reviewer family) is ARIS’s core invariant: an LLM auditing its own output is no audit.
🌐 ARIS-Homepage — fact-checked academic homepage from CV
The only personal-site generator that fact-checks your CV before publishing.
A new skill in this repo: /homepage-generator turns your CV (.docx / .pdf / .txt) into a polished single-file academic homepage. A DBLP audit runs on every render unless you pass --no-audit — a wrong venue or year, or a Best Paper / Spotlight / Oral / Outstanding badge asserted without a link to the paper, blocks the render until corrected or explicitly overridden. (It does not check author lists.)
Live demo: wanshuiyin.github.io — generated by this skill from a CV + the maintainer’s previous manual page as editorial reference. Preview strip is near the top of this README.
Quick start
There is no aris-homepage executable — this repo ships no installer, so call the script directly. The path is relative to where you are standing, which changes at the cd:
python tools/aris_homepage.py init --from-cv ./cv.pdf --out ./site
cd ./site
# Calling agent fills .aris-homepage/extraction.json per EXTRACTION_HANDOFF.md
python ../tools/aris_homepage.py finalize
$EDITOR profile.yml # tweak editorial choices
python ../tools/aris_homepage.py render --persona theory-minimal
🪟 On Windows? Use \ in the paths above, and mind the Windows-only traps — the python3 alias stub, silent .py file association, non-ASCII pdftotext paths, SSL certificates. Full walkthrough: WINDOWS_en.md (中文).
Output: index.html + audit-report.md. Drop the HTML on GitHub Pages, S3, university ~user/public_html/, or attach to email — no build server. Minimum runtime is just Python + a calling LLM agent; Codex MCP optional for adversarial cross-model review; Gemini multimodal optional for visual critique.
How it works
ARIS-Homepage Pipeline
📄 CV (.docx/.pdf/.txt) 🌐 Manual Homepage URL 🖼 Assets Dir
factual source editorial (optional) visual (opt.)
│ │ │
▼ │ │
┌──────────┐ │ │
│ init │ │ │
│ extract │ │ │
│ CV→text │ │ │
└─────┬────┘ │ │
▼ ▼ ▼
┌─────────────────────────────────────────────────────────────────┐
│ 🤖 Calling LLM agent reads EXTRACTION_HANDOFF.md + │
│ optional manual-homepage URL + assets dir as context │
│ → writes .aris-homepage/extraction.json │
└─────────────────────────┬───────────────────────────────────────┘
▼
┌──────────┐
│ finalize │
└─────┬────┘
▼
┌─────────────────────────────────────────────────────────────────┐
│ ✋ Editable source files (truth lives here, edit in IDE): │
│ profile.yml · publications.bib · bio.md · news.md │
│ EXTRACTION_REVIEW.md (review LLM uncertain extractions) │
└─────────────────────────┬───────────────────────────────────────┘
▼
┌────────────────────────┐
│ render │
│ --persona │
│ theory-minimal │
└───────────┬────────────┘
│
┌───────────────┼───────────────┐
▼ ▼ ▼
┌──────────┐ ┌──────────┐ ┌──────────────┐
│ Layer-1 │ │ Layer-2 │ │ Layer-2 │
│ DBLP │ │ Codex MCP│ │ Gemini │
│ fact-chk │ │ adv-rev │ │ visual │
│ │ │ (opt.) │ │ critique │
│ (default)│ │ │ │ (opt.) │
└─────┬────┘ └──────────┘ └──────────────┘
│
▼
┌──────────────┐
│ index.html + │
│ audit-report │ ──▶ 🚀 Deploy: GitHub Pages · S3 · email · anywhere
│ .md │
└──────────────┘
Typical flow (7 steps, ~5 minutes):
1. python tools/aris_homepage.py init --from-cv ./cv.pdf --out ./site
2. (calling agent) read .aris-homepage/EXTRACTION_HANDOFF.md
→ fill .aris-homepage/extraction.json
3. cd ./site && python ../tools/aris_homepage.py finalize
4. $EDITOR profile.yml publications.bib bio.md news.md
5. python ../tools/aris_homepage.py check --strict # fact-check only
6. python ../tools/aris_homepage.py render --persona theory-minimal
7. inspect audit-report.md; fix → re-render OR --override-all
Minimum runtime: Python + a calling LLM agent.
Codex MCP optional (cross-model adversarial review).
Gemini optional (multimodal visual critique).
🤝 Contributing
One person can only cover so much. The hope is that many hands make this collection more complete.
Full contribution guide: CONTRIBUTING.md (English · 中文) — covers ARIS workflow invocation, strict style guide (headings / math / tables / callouts / personal-info banlist), and PR checklist.
TL;DR: use the /interview-cheatsheet + /render-html workflow to generate, then open a PR. Both skills enforce a cross-model codex GPT-5.5 xhigh review gate (math / code / citation / render fidelity), so anything merged via PR has a baseline quality floor. Skill source and tools/render_html.py are bundled in this repo so you can fork & extend.
Honest disclaimer: across the existing tutorials, the HTML structural foundations (math, code, tables, callouts, TOC, responsive layout) are solid. But the very latest frontier work in any given topic (e.g., methods released in late 2025, niche subfield updates) likely is not fully covered. If you spot something outdated or wrong, PRs and issues are equally welcome — let’s keep this resource alive together.
Shared community with the main ARIS repo — the same WeChat group covers ARIS skill workflows + this tutorial collection. Join to discuss interview prep, request new cheat-sheet topics, or share corrections / contributions:
Community-built projects derived from this collection (MIT license — attribution-preserving reuse welcome):
Built something on top of these tutorials? Open an issue and we’ll list it here.
🌟 What is ARIS — A Quick Pitch
ARIS — Auto Research in Sleep is one of the most-watched AI research agent skill platforms of 2025-2026. The /interview-cheatsheet + /render-html skills that produced this repo are 2 out of ARIS’s 74+ skills.
· · · · ·
- ⭐ ~10k GitHub stars — top-trending AI agent repo
- 🥇 HuggingFace Daily Papers #1 — top of the day, paper arXiv:2605.03042
- 🏆 AI Digital Crew · Project of the Day (2026.03.14)
- 📰 Featured on PaperWeekly + VoltAgent/awesome-agent-skills
- 🛠️ 74+ research skills — full lifecycle from idea exploration → experiments → papers → rebuttals → talk slides
- 🌐 7+ platforms supported — Claude Code · Codex CLI · Cursor · Trae · Antigravity · GitHub Copilot CLI · OpenClaw
- 🔧 ARIS-Code standalone CLI — multi-provider runtime, no Claude Code dependency required
Core methodology: cross-model adversarial review — executor and reviewer must come from different model families (Claude × GPT-5.5 xhigh × Gemini), so no LLM ever judges its own output. This protocol carries directly into interview cheat sheet generation: every formula, code block, and citation in every tutorial passes an independent audit (see each .review.json audit trail).
👉 ARIS main repo: https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep
📖 Citing ARIS
If this collection — or any cheat sheet here — helped you in your interview prep / research / paper, please consider citing the underlying ARIS methodology paper:
@article{yang2026aris,
title={ARIS: Autonomous Research via Adversarial Multi-Agent Collaboration},
author={Yang, Ruofeng and Li, Yongcan and Li, Shuai},
journal={arXiv preprint arXiv:2605.03042},
year={2026}
}
Every tutorial in this repo was generated end-to-end by the ARIS /interview-cheatsheet + /render-html workflow with cross-model adversarial review (Claude × GPT-5.5 xhigh × Gemini). The citation supports the methodology behind the workflow, not just this collection.
License
MIT — use, modify, share, fork freely. Hope this helps your job search. 💪
评论