INDEX

Explanations

phrases related to content creation, potentially including scripts, recipes, and copyright information

oai_token-act-pair · gpt-3.5-turbo

references to specific entities or characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 responsible

-0.69

 Calder

-0.67

 ACTION

-0.57

 Clover

-0.57

DERR

-0.57

 necessary

-0.56

 Surge

-0.56

 defunct

-0.55

 progressing

-0.55

roit

-0.55

POSITIVE LOGITS

atoon

0.99

pter

0.96

ffield

0.91

cer

0.89

udos

0.80

cific

0.79

raper

0.76

kered

0.75

hett

0.74

glers

0.73

Activations Density 0.055%

phrases related to content creation, potentially including scripts, recipes, and copyright information

references to specific entities or characters

No Comments

No Known Activations