INDEX

Explanations

proper nouns related to political figures, specifically Barack Obama

oai_token-act-pair · gpt-3.5-turbo

references to acknowledgement or recognition

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Debor

-0.75

)].

-0.69

 conflic

-0.68

senal

-0.67

Tib

-0.66

 eleph

-0.66

 Seym

-0.65

 warr

-0.64

 livest

-0.63

 dough

-0.62

POSITIVE LOGITS

enzie

1.19

ack

1.03

intosh

1.02

ACK

0.97

acks

0.96

acked

0.95

ademic

0.93

luster

0.89

nesses

0.86

les

0.84

Activations Density 0.011%

proper nouns related to political figures, specifically Barack Obama

references to acknowledgement or recognition

No Comments

No Known Activations