INDEX

Explanations

mentions of "physician"

oai_token-act-pair · gpt-3.5-turbo

references to physicians and their assessments of health

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

urat

-0.87

yip

-0.85

yrinth

-0.77

skirts

-0.75

footed

-0.74

eworld

-0.74

runners

-0.72

nings

-0.72

eat

-0.72

ptions

-0.70

POSITIVE LOGITS

 practitioner

1.00

 physician

1.00

 physicians

0.84

 examiner

0.84

 prescribing

0.83

 prescriptions

0.82

 psychiatrist

0.81

 doctor

0.81

 prescription

0.80

 specializing

0.79

Activations Density 0.015%

mentions of "physician"

references to physicians and their assessments of health

No Comments

No Known Activations