INDEX

Explanations

terms related to action genres, such as action movies and action-packed experiences

oai_token-act-pair · gpt-3.5-turbo

references to action-related content in various contexts

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

gling

-0.76

thening

-0.73

YL

-0.70

mbuds

-0.67

gins

-0.62

owship

-0.62

wn

-0.61

yth

-0.61

DAQ

-0.61

heed

-0.61

POSITIVE LOGITS

 Replay

0.80

tec

0.78

ãĤ¿

0.77

quit

0.74

ivated

0.73

bucks

0.71

ual

0.71

="/

0.68

ional

0.68

 hero

0.68

Activations Density 0.024%

terms related to action genres, such as action movies and action-packed experiences

references to action-related content in various contexts

No Comments

No Known Activations