INDEX

Explanations

references to different pieces of media such as movies, albums, and interviews

oai_token-act-pair · gpt-3.5-turbo

references to specific films and their production details

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

 administr

-0.61

jri

-0.59

 confer

-0.59

etooth

-0.57

nel

-0.56

 disag

-0.52

 jurisd

-0.52

 advant

-0.52

isance

-0.50

 wealthier

-0.49

POSITIVE LOGITS

 Ghostbusters

1.01

 screenplay

0.97

 starring

0.95

 Deadpool

0.95

 Godzilla

0.93

Batman

0.91

episode

0.91

 Daredevil

0.91

 Comedy

0.91

HBO

0.90

Activations Density 0.791%

references to different pieces of media such as movies, albums, and interviews

references to specific films and their production details

No Comments

No Known Activations