INDEX

Explanations

Elizabeth Warren

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Warren," particularly focusing on Warren Buffett and Elizabeth Warren

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 spat

-0.80

pn

-0.76

ip

-0.74

 cookies

-0.73

eth

-0.72

DH

-0.72

ãĥĥãĥī

-0.68

dh

-0.66

 salty

-0.65

ic

-0.65

POSITIVE LOGITS

 Warren

3.94

Warren

3.04

 Buffett

1.40

 Berkshire

1.26

 Franklin

1.18

 Woodward

1.14

 Rockefeller

1.13

 Russell

1.11

 McConnell

1.10

 Ellison

1.09

Activations Density 0.014%

**Elizabeth Warren**

mentions of the name "Warren," particularly focusing on Warren Buffett and Elizabeth Warren

No Comments

No Known Activations

Elizabeth Warren