INDEX

Explanations

phrases indicating an action of looking at or reviewing something

oai_token-act-pair · gpt-3.5-turbo

phrases that suggest examining or discussing various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

cell

-0.72

Church

-0.70

trap

-0.69

widget

-0.68

CI

-0.68

theless

-0.66

operated

-0.66

isSpecialOrderable

-0.66

CH

-0.66

soever

-0.66

POSITIVE LOGITS

 whats

0.80

how

0.71

 puberty

0.69

ĸļ

0.69

mosp

0.65

abase

0.65

 least

0.63

 finer

0.63

the

0.63

 angles

0.63

Activations Density 0.032%

phrases indicating an action of looking at or reviewing something

phrases that suggest examining or discussing various subjects

No Comments

No Known Activations