INDEX

Explanations

phrases related to activities or events that are not easily visible or known to the public, typically happening backstage or off-camera

oai_token-act-pair · gpt-3.5-turbo

references to "behind the scenes" activities or contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

heit

-0.82

rill

-0.78

mington

-0.78

arcity

-0.78

urdue

-0.74

haps

-0.72

gling

-0.71

apa

-0.71

liam

-0.70

rab

-0.69

POSITIVE LOGITS

 filming

0.74

spection

0.71

 reconnaissance

0.71

 performer

0.69

 workings

0.67

 scenes

0.67

 talents

0.66

 coaching

0.66

 machinery

0.65

 camera

0.64

Activations Density 0.038%

phrases related to activities or events that are not easily visible or known to the public, typically happening backstage or off-camera

references to "behind the scenes" activities or contexts

No Comments

No Known Activations