Research Scientist/Engineer, Honesty at Anthropic

San Francisco, California, United States

Apply Now

Not SpecifiedCompensation

Senior (5 to 8 years)Experience Level

Full TimeJob Type

UnknownVisa

Artificial Intelligence, Machine LearningIndustries

Requirements

MS/PhD in Computer Science, ML, or related field
Strong programming skills in Python
Industry experience with language model finetuning and classifier training
Proficiency in experimental design and statistical analysis for measuring improvements in calibration and accuracy
Care about AI safety and the accuracy and honesty of both current and future AI systems
Experience in data science or the creation and curation of datasets for finetuning LLMs
An understanding of various metrics of uncertainty, calibration, and truthfulness in model outputs

Responsibilities

Design and implement novel data curation pipelines to identify, verify, and filter training data for accuracy given the model’s knowledge
Develop specialized classifiers to detect potential hallucinations or miscalibrated claims made by the model
Create and maintain comprehensive honesty benchmarks and evaluation frameworks
Implement search and retrieval-augmented generation (RAG) systems to ground model outputs in verified information
Design and deploy human feedback collection specifically for identifying and correcting miscalibrated responses
Design and implement prompting pipelines to generate data that improves model accuracy and honesty
Develop and test novel RL environments that reward truthful outputs and penalize fabricated claims
Create tools to help human evaluators efficiently assess model outputs for accuracy

Skills

Python

LLM Finetuning

Classifier Training

Experimental Design

Statistical Analysis

Data Curation

Dataset Creation

AI Safety

Calibration Metrics

Uncertainty Metrics

Truthfulness Metrics

Anthropic

Develops reliable and interpretable AI systems

About Anthropic

Anthropic focuses on creating reliable and interpretable AI systems. Its main product, Claude, serves as an AI assistant that can manage tasks for clients across various industries. Claude utilizes advanced techniques in natural language processing, reinforcement learning, and code generation to perform its functions effectively. What sets Anthropic apart from its competitors is its emphasis on making AI systems that are not only powerful but also understandable and controllable by users. The company's goal is to enhance operational efficiency and improve decision-making for its clients through the deployment and licensing of its AI technologies.

San Francisco, CaliforniaHeadquarters

2021Year Founded

$11,482.1MTotal Funding

GROWTH_EQUITY_VCCompany Stage

Enterprise Software, AI & Machine LearningIndustries

1,001-5,000Employees

Benefits

Flexible Work Hours

Paid Vacation

Parental Leave

Hybrid Work Options

Company Equity

Risks

Ongoing lawsuit with Concord Music Group could lead to financial liabilities.

Technological lag behind competitors like OpenAI may impact market position.

Reliance on substantial funding rounds may indicate financial instability.

Differentiation

Anthropic focuses on AI safety, contrasting with competitors' commercial priorities.

Claude, Anthropic's AI assistant, is designed for tasks of any scale.

Partnerships with tech giants like Panasonic and Amazon enhance Anthropic's strategic positioning.

Upsides

Anthropic's $60 billion valuation reflects strong investor confidence and growth potential.

Collaborations like the Umi app with Panasonic tap into the growing wellness AI market.

Focus on AI safety aligns with increasing industry emphasis on ethical AI development.

Land your dream remote job 3x faster with AI

Try Jobo Free