INDEX

Explanations

mentions of specific TV networks and series

oai_token-act-pair · gpt-3.5-turbo

mentions of HBO and associated programming

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

dyl

-0.74

 Fiorina

-0.74

 Mata

-0.73

Syl

-0.66

Riy

-0.66

 Jere

-0.65

Alz

-0.65

etooth

-0.65

quez

-0.63

 Elise

-0.63

POSITIVE LOGITS

 Films

1.25

Universal

1.16

 Networks

1.15

 Television

1.13

 Studios

1.08

 sitcom

1.06

 Presents

1.03

 Movies

1.03

 Productions

0.98

 Premiere

0.95

Activations Density 0.095%

mentions of specific TV networks and series

mentions of HBO and associated programming

No Comments

No Known Activations