INDEX

Explanations

references or allusions to specific entities or concepts

oai_token-act-pair · gpt-3.5-turbo

phrases indicating homage or tribute to other works or influences

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

DH

-0.82

ienne

-0.77

eret

-0.68

ctl

-0.67

ynski

-0.62

een

-0.62

dk

-0.62

efficients

-0.61

tails

-0.61

confidence

-0.61

POSITIVE LOGITS

rium

0.78

 commemorate

0.77

 appease

0.71

 conserve

0.70

 Parables

0.69

 Uran

0.67

 antiquity

0.67

 Pagan

0.67

 populate

0.66

 Humanity

0.66

Activations Density 0.145%

references or allusions to specific entities or concepts

phrases indicating homage or tribute to other works or influences

No Comments

No Known Activations