INDEX

Explanations

phrases indicating the presence or existence of something within a particular context

oai_token-act-pair · gpt-3.5-turbo

instances of the word "contain" and its variations, indicating the presence of certain elements or substances

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

doms

-0.85

 Sabha

-0.79

 Seym

-0.71

urai

-0.69

 Airways

-0.63

yrim

-0.63

 athlet

-0.62

straight

-0.61

boarding

-0.61

hon

-0.60

POSITIVE LOGITS

ãĤ£

0.75

encies

0.75

ãĤ¼ãĤ¦ãĤ¹

0.74

ttes

0.71

ãĥ¯ãĥ³

0.70

 sufficient

0.70

 contents

0.70

 explosives

0.69

 lots

0.68

nces

0.68

Activations Density 0.039%

phrases indicating the presence or existence of something within a particular context

instances of the word "contain" and its variations, indicating the presence of certain elements or substances

No Comments

No Known Activations