INDEX

Explanations

acronyms that refer to different technologies or concepts

oai_token-act-pair · gpt-3.5-turbo

abbreviations, acronyms, or symbols related to various specialized topics or industries

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

fried

-0.74

 UNIVERS

-0.71

 Newsp

-0.70

airo

-0.69

fif

-0.67

pull

-0.67

 WATCHED

-0.67

ãĤ©

-0.66

Lud

-0.65

 Printed

-0.65

POSITIVE LOGITS

 inhibitors

0.81

 inhibitor

0.80

Gs

0.78

 activ

0.76

 activation

0.74

Ds

0.72

+.

0.72

 attainment

0.71

 buckets

0.70

ACY

0.69

Activations Density 0.146%

acronyms that refer to different technologies or concepts

abbreviations, acronyms, or symbols related to various specialized topics or industries

No Comments

No Known Activations