INDEX

Explanations

references to "Pew Research" or "Pew Research Center" in text

oai_token-act-pair · gpt-3.5-turbo

mentions of the Pew Research Center

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Runner

-0.68

ACTION

-0.66

 Centauri

-0.62

 DRAG

-0.61

 fluids

-0.59

 seriousness

-0.59

CONCLUS

-0.58

 waking

-0.58

 severity

-0.58

 ultras

-0.58

POSITIVE LOGITS

aukee

1.15

Die

1.00

Pew

0.99

die

0.89

sburg

0.85

pie

0.82

auld

0.80

itzer

0.79

sta

0.79

ersen

0.78

Activations Density 0.006%

references to "Pew Research" or "Pew Research Center" in text

mentions of the Pew Research Center

No Comments

No Known Activations