INDEX

Explanations

mentions of the word "Comics."

oai_token-act-pair · gpt-3.5-turbo

mentions of comic books and comic-related content

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

ntil

-0.87

holm

-0.81

wreck

-0.73

ĪĴ

-0.71

beam

-0.70

tenance

-0.70

unci

-0.69

gren

-0.69

ŃĶ

-0.68

ysis

-0.67

POSITIVE LOGITS

 Comics

1.11

 comics

0.93

 superheroes

0.85

 superhero

0.81

 Digest

0.80

 adaptations

0.79

 adaptation

0.77

 Presents

0.77

 sans

0.76

 publishers

0.75

Activations Density 0.014%

mentions of the word "Comics."

mentions of comic books and comic-related content

No Comments

No Known Activations