INDEX

Explanations

scenes described or depicted in a document

oai_token-act-pair · gpt-3.5-turbo

references to specific scenes in movies or narratives

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

orem

-0.92

iott

-0.83

ikarp

-0.74

ileen

-0.73

gemony

-0.73

achev

-0.70

thood

-0.70

sonian

-0.70

milo

-0.68

dinand

-0.68

POSITIVE LOGITS

 unfolding

1.05

 involving

1.05

Dialogue

1.04

 scenes

1.02

 depicting

1.01

 Scenes

0.97

 unfold

0.96

 filmed

0.94

Capture

0.93

 Scene

0.89

Activations Density 0.043%

scenes described or depicted in a document

references to specific scenes in movies or narratives

No Comments

No Known Activations