Janghoon Lee

English/한국어

References

Papers on the path into the generative and eval stack I work on. Each line is what that work is for here, not an abstract.

  • Attention Is All You Need2017 · Vaswani et al.Introduces the Transformer. The attention backbone under every modern LLM and most generative stacks since.
  • GPT-22019 · Radford et al.Shows unsupervised next-token training at scale yields transferable task behavior. Template for “train once, adapt with prompts.”
  • Scaling Laws2020 · Kaplan et al.Predictable loss vs compute, data, and parameters. The reason bigger models and careful budget allocation beat folklore.
  • DDPM2020 · Ho, Jain, AbbeelDenoising diffusion as a practical generative process. Foundation for image (and later video) diffusion systems.
  • CLIP2021 · Radford et al.Joint image-text embeddings from contrastive training. How text conditions modern image models and how we score alignment.
  • Latent Diffusion2022 · Rombach et al.Runs diffusion in a compressed latent space so high-res image generation is affordable. Ancestor of Stable Diffusion-class stacks we tested and rejected for natural-prompt anatomy.
  • Video Diffusion2022 · Ho et al.Extends diffusion to video. Establishes the denoising-over-time framing later productized at frontier scale.
  • EnCodec2022 · Défossez et al.Neural audio codec: speech and music as discrete tokens. Bridge from waveforms into language-model-style generation.
  • DiT2022 · Peebles, XieDiffusion Transformers: replace the U-Net with a transformer over latent patches. Architecture family behind frontier image generators and ACE Step’s DiT path.
  • Flow Matching2023 · Lipman et al.Trains continuous normalizing flows by regressing a vector field. Cleaner, often faster alternative path to diffusion-style generation.
  • FrugalGPT2023 · Chen, Zaharia, ZouPrompt adaptation, approximation, and LLM cascades to cut API cost. Early framing for cheap-vs-strong routing that Redrob Eval and Console sit downstream of.
  • RouteLLM2024 · Ong et al.Learns routers from preference data to send easy queries to a weak model and hard ones to a strong model. Framework we tried in Eval before dropping routing at ~25% label match.
  • LiteLLM2024 · BerriAIOpen-source multi-provider LLM gateway and routing toolkit. One of the RouteLLM-adjacent stacks exercised in the Eval routing failure.
  • APG2024 · Sadat et al.Adaptive Projected Guidance: fixes CFG oversaturation by projecting guidance updates. Why Tune Pro uses APG instead of plain DiT CFG on KSampler.
  • CosyVoice 22024 · Du et al.Streaming, low-latency speech synthesis with strong zero-shot voice cloning. Reference point for speech models in the Redrob Speech direction.
  • FLUX.1 Kontext2025 · Black Forest Labs et al.Flow-matching generation and editing from Black Forest Labs. Lineage for the FLUX.2 [klein] 4B photoreal candidate we tested and rejected on typography.
  • GEPA2025 · Agrawal et al.Genetic-Pareto reflective prompt evolution. How Redrob Eval optimizes the cheap path before dual-eval on IN22-Gen, IndicGLUE, and GSM8K.
  • Qwen-Image2025 · Wu et al.Image foundation model with strong text rendering and native editing. Heavy path in Redrob Image: custom checkpoints and edit-first product work.
  • Z-Image2025 · Tongyi-MAI / Z-Image TeamEfficient single-stream DiT image model; Z-Image-Turbo is the distilled 8-step variant. Fast path for quick and easy Redrob Image requests.
  • Sarvam2026 · Sarvam AIOpen Indic-focused language models (30B / 105B and earlier Sarvam-1/M). Open backend family behind the routed Indic text stack in Redrob Console.
  • ACE-Step 1.52026 · Gong et al.Open-source music foundation model: LM planner plus DiT, local inference, LoRA personalization. Weights and Comfy graph behind Redrob Tune Fast/Pro.