🏗️ Building on HF

Kshitij Thakkar PRO

kshitijthakkar

·

AI & ML interests

Building the evaluation and observability layer for AI. Creator of TraceVerse—turning real-world LLM interactions into datasets, benchmarks, and cost-efficient model insights.

Recent Activity

updated a bucket 2 days ago

guardian-tails/sentinel-kickoff

updated a bucket 2 days ago

guardian-tails/sentinel-chiku-inu

published a bucket 2 days ago

guardian-tails/sentinel-chiku-inu

View all activity

Organizations

New activity in rl-llm-wiki/knowledge-base 25 days ago

source: arxiv:2607.05904 - More Convincing, Not More Correct (self-play reward hacking of reference-free LLM judges)

#616 opened 25 days ago by

source: arxiv:2412.16339 — Deliberative Alignment (Reasoning Enables Safer LMs)

#595 opened 28 days ago by

New activity in rl-llm-wiki/knowledge-base 27 days ago

topic: judging-bias-and-contamination — deepen + add citations

#593 opened 29 days ago by

topic: llm-as-judge — deepen + add citations

#583 opened 29 days ago by

topic: hallucination-and-abstention developing->comprehensive (o1-card frontier weave + monitor-as-verifier; 16 anchors verified)

#596 opened 27 days ago by

source: arxiv:2203.07814 — Competition-Level Code Generation with AlphaCode

#584 opened 29 days ago by

source: arxiv:2005.01643 — Offline RL: Tutorial, Review, and Perspectives on Open Problems

#592 opened 29 days ago by

source: arxiv:1910.00177 — Advantage-Weighted Regression (AWR)

#591 opened 29 days ago by

source: arxiv:1609.05473 — SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient

#585 opened 29 days ago by

source: arxiv:2006.04779 — Conservative Q-Learning for Offline Reinforcement Learning

#590 opened 29 days ago by

source: arxiv:2106.01345 — Decision Transformer

#581 opened 29 days ago by

source: arxiv:2004.07219 — D4RL: Datasets for Deep Data-Driven Reinforcement Learning

#586 opened 29 days ago by

source: arxiv:1707.01495 — Hindsight Experience Replay

#588 opened 29 days ago by

New activity in rl-llm-wiki/knowledge-base about 1 month ago

topic: hallucination-and-abstention under RL (new; phenomena-and-failure-modes)

#503 opened about 1 month ago by

source: arxiv:2401.01335 — SPIN (Self-Play Fine-Tuning)

#500 opened about 1 month ago by

source: arxiv:2503.02846 - Mask-DPO (fine-grained factuality alignment, ICLR 2025)

#502 opened about 1 month ago by

topic: adversarial-robustness-and-jailbreaks — add GPT-4 fine-tuning attack + RLHF data-poisoning surface; developing → comprehensive

#428 opened about 1 month ago by

source: arxiv:2311.08045 — Adversarial Preference Optimization (APO)

#427 opened about 1 month ago by

source: arxiv:2310.04921 — Crystal: Introspective Reasoners Reinforced with Self-Feedback

#424 opened about 1 month ago by

source: arxiv:2308.06385 — ZYN: Zero-Shot Reward Models with Yes-No Questions for RLAIF

#417 opened about 1 month ago by