INDEX

Explanations

mentions of the word "franchise."

oai_token-act-pair · gpt-3.5-turbo

references to movie or entertainment franchises

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

board

-0.66

Org

-0.66

ead

-0.65

lap

-0.65

 Library

-0.64

agher

-0.64

ibrary

-0.64

 Trance

-0.63

 Marketable

-0.62

Latest

-0.61

POSITIVE LOGITS

 franchise

1.07

 franchises

0.96

 Franchise

0.90

eers

0.87

arily

0.83

 franch

0.82

rats

0.80

terness

0.73

ovies

0.71

adelphia

0.70

Activations Density 0.008%

mentions of the word "franchise."

references to movie or entertainment franchises

No Comments

No Known Activations