---
schema: you-md/v1
name: Priya Sharma
username: youmdqaa45181da2c
generated_at: 2026-04-16T22:55:37.583Z
---

# Priya Sharma
ML engineer @ Anthropic. Alignment researcher.
*London, UK*

## About

Research engineer focused on RLHF and alignment. ML engineer at Anthropic working on reward modeling and the social dynamics of evaluation. I think in probability distributions and communicate in analogies.
Published 12 papers on reward modeling. Top 1% cited in ML safety. Weekend ceramicist — turns out throwing pots and shaping reward landscapes have more in common than you'd think.
I care about rigorous evaluation, open science, and making sure the systems we ship today are the ones we want to live with tomorrow.

## Projects

- **Reward Landscapes** -- Novel approach to multi-objective reward modeling. Investigating how reward models generalize when objectives conflict, and what that tells us about preference aggregation in fine-tuning. (Lead researcher)
- **EvalKit** -- Open-source LLM evaluation framework. Composable eval suites with reproducible benchmarks and structured failure analysis. (Creator)
- **Calibration in Critique** -- Ongoing research on how well LLM critics know what they don't know, and how miscalibration propagates into downstream training signal. (Co-author)
- **Pottery Studio** -- Wheel-throwing on weekends. Currently obsessed with celadon glazes and reduction firing. (Apprentice)

## Values

- Rigorous thinking — claims should survive their own ablation
- Open science — share negative results, share evals, share weights when you can
- Calibrated uncertainty — say what you actually know
- Empirical humility — the model is almost always smarter than your intuition
- Make it safe — alignment is the work, not the side quest

## Agent Preferences

Tone: precise, curious, grounded
Avoid: hype language, unsubstantiated claims, anthropomorphizing models, "obviously"

## Links

- Google Scholar: https://scholar.google.com/citations?user=priya
- GitHub: https://github.com/priya
- Anthropic: https://www.anthropic.com
- Personal site: https://priya.dev
- EvalKit: https://github.com/priya/evalkit

## Voice

Technical but approachable. Loves analogies — usually drawn from physics, pottery, or cooking. Sentences carry weight; nothing is filler.

Common patterns:
- Open with the precise question being asked
- Use analogies to make the formal claim land
- Footnotes for the rigorous bits
- Always end with "what this implies for evaluation"
- Lowercase "i" in informal writing, full caps for variables and acronyms

---

> **For agents**: this is a you-md/v1 identity context protocol.
> Quick context: check identity.bio.short, now.focus, and preferences.agent.
> For writing help: check voice section and preferences.writing.
> Full structured data: see you.json.