INDEX

Explanations

mentions of cartoon characters or TV shows

oai_token-act-pair · gpt-3.5-turbo

mentions of cartoons

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

HI

-0.71

govern

-0.69

 sclerosis

-0.69

acia

-0.69

forces

-0.68

alez

-0.67

FUL

-0.66

ttp

-0.64

utherford

-0.64

vae

-0.64

POSITIVE LOGITS

 cartoons

1.08

ishly

1.05

 cartoon

0.96

 frog

0.91

 caric

0.89

 sketches

0.87

ists

0.85

 Cartoon

0.84

 depictions

0.84

ist

0.83

Activations Density 0.018%

mentions of cartoon characters or TV shows

mentions of cartoons

No Comments

No Known Activations