INDEX

Explanations

words related to agreement or consensus

oai_token-act-pair · gpt-3.5-turbo

words related to opinions and concurrent events or scenarios

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

Kov

-0.71

handler

-0.68

otechnology

-0.65

 Container

-0.64

 Jess

-0.64

endix

-0.63

 Tate

-0.63

erion

-0.63

 Tiff

-0.63

 dumps

-0.62

POSITIVE LOGITS

rences

1.15

interstitial

0.96

olars

0.96

urring

0.86

urred

0.82

SHIP

0.79

ROR

0.77

ienced

0.76

urs

0.76

VIDIA

0.76

Activations Density 0.012%

words related to agreement or consensus

words related to opinions and concurrent events or scenarios

No Comments

No Known Activations