INDEX

Explanations

phrases containing the word "Jude."

oai_token-act-pair · gpt-3.5-turbo

references to the name "Jude."

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

ickr

-0.88

asks

-0.86

ending

-0.84

rings

-0.79

awaru

-0.77

tackle

-0.75

isting

-0.75

fight

-0.74

quished

-0.72

bnb

-0.71

POSITIVE LOGITS

 Galile

1.17

 Jude

1.11

Whe

0.79

 Caesar

0.75

 Cornwall

0.73

 Kirin

0.69

 Kush

0.67

 Crus

0.66

 Jonah

0.65

 Kane

0.65

Activations Density 0.015%

phrases containing the word "Jude."

references to the name "Jude."

No Comments

No Known Activations