INDEX

Explanations

references to movies, books, and characters from various franchises or series

oai_token-act-pair · gpt-3.5-turbo

references to popular television series and their associated characters or themes

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

 seiz

-0.71

lb

-0.69

 reconc

-0.68

HRC

-0.66

Fax

-0.65

 disag

-0.62

tenance

-0.62

 vulner

-0.62

 lawy

-0.61

 damages

-0.61

POSITIVE LOGITS

 trilogy

1.37

 mythology

1.23

 novels

1.21

 sequels

1.18

 franchise

1.17

 franchises

1.17

 lore

1.10

 classics

1.09

 series

1.09

 universes

1.09

Activations Density 0.409%

references to movies, books, and characters from various franchises or series

references to popular television series and their associated characters or themes

No Comments

No Known Activations