INDEX

Explanations

terms related to government bureaus or organizations

oai_token-act-pair · gpt-3.5-turbo

mentions of various governmental or organizational bureaus

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

terday

-0.71

isable

-0.70

 Vader

-0.68

IFE

-0.68

wcsstore

-0.67

enger

-0.65

isks

-0.64

orem

-0.64

 wart

-0.64

 lear

-0.63

POSITIVE LOGITS

ureau

1.04

oreal

0.95

cr

0.86

zag

0.76

 Reloaded

0.75

ration

0.73

xia

0.72

cles

0.70

omial

0.70

vernment

0.69

Activations Density 0.022%

terms related to government bureaus or organizations

mentions of various governmental or organizational bureaus

No Comments

No Known Activations