INDEX

Explanations

phrases related to a specific keyword "SM" followed by a number (e.g., SM 9, SM 10)

oai_token-act-pair · gpt-3.5-turbo

references to a specific series of media or entertainment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ved

-0.78

 Cerberus

-0.70

 Gamergate

-0.70

 Kurd

-0.68

bart

-0.68

 Greenwald

-0.68

ãĤ¡

-0.67

Nou

-0.67

 WikiLeaks

-0.66

reon

-0.66

POSITIVE LOGITS

SM

0.99

essage

0.96

ITH

0.93

ailability

0.87

ASH

0.86

 mosqu

0.85

ategory

0.85

oke

0.85

ugg

0.84

 cryst

0.84

Activations Density 0.003%

phrases related to a specific keyword "SM" followed by a number (e.g., SM 9, SM 10)

references to a specific series of media or entertainment

No Comments

No Known Activations