INDEX

Explanations

opinions or preferences expressed in survey data

oai_token-act-pair · gpt-3.5-turbo

references to public opinion and trust in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

inis

-0.92

GV

-0.73

CVE

-0.64

uez

-0.64

tl

-0.63

esis

-0.63

issance

-0.62

ortment

-0.61

pter

-0.60

CVE

-0.60

POSITIVE LOGITS

 electing

0.80

 legalizing

0.79

 voting

0.78

 abortion

0.77

 disapprove

0.76

 unfavorable

0.75

 suicide

0.74

 subscribing

0.74

 independence

0.73

ecided

0.72

Activations Density 0.179%

opinions or preferences expressed in survey data

references to public opinion and trust in various contexts

No Comments

No Known Activations