Hi, I'm Jaydeep

Top 100 at the HuggingFace × Meta OpenEnv Hackathon, Bangalore. Built an RL environment where a small monitor model learns to catch a frontier attacker (DeepSeek-R1) attempting hidden harmful behaviors across 9 enterprise scenarios. Every reward is a plain Python check, with no LLM-as-judge anywhere in the loop. After 10 reward design iterations, a simple linear reward beat the more elaborate composed one. Trained a Qwen-2.5-1.5B LoRA monitor from random performance (AUROC 0.500) to strong detection (0.893, Recall 0.88, FPR 0.12), closing about 40% of the gap to Gemini-2.5-Pro at under 0.1% of its per-call cost.

Continued pre-training of SmolLM-135M on 138 arXiv ML papers (2024–2026) using QLoRA (rank 32) via Unsloth. Trained on an RTX 4090 in about 14 minutes. On held-out papers: −20.1% perplexity (22.97 to 18.36), +19.7% ROUGE-1, +25.4% ROUGE-L, +37.5% BLEU. Two takeaways: LoRA outperforms full fine-tuning on small datasets thanks to its regularization, and rank saturates at r≥16, which suggests the data (not the rank) is the bottleneck. Adapter (9.7M trainable params, 6.77% of total) published to HuggingFace.

Two diffusion language models, both built from scratch. ModernBERT (~150M): pretrained as a masked diffusion LM on Project Gutenberg (6.4M chunks, 20 hours on an RTX 4090), then SFT'd on Open-Orca (~4.2M Q&A pairs, instruction-token-only loss). TinyStories (45M): full architecture written from scratch and trained for 60K steps, with confidence-based iterative denoising at generation time (128 diffusion steps). Loss dropped sharply around step 25K, the point where structure starts to emerge. All checkpoints published to HuggingFace.

End-to-end system for evaluating government tenders. The pipeline pulls criteria from the document with a DeepSeek LLM, retrieves matching evidence with sentence-transformers, and writes an explainable verdict with cited source clauses. OCR uses a three-tier fallback: PyMuPDF first, Tesseract next, DeepSeek Vision when confidence drops below 65%. Borderline verdicts (confidence 0.55 to 0.80) route to a human review queue with a full audit trail. Covered by a 43-check smoke test suite.