INDEX

Explanations

words associated with actions or events happening in the past

oai_token-act-pair · gpt-3.5-turbo

punctuation and specific characters in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

agar

-1.03

sag

-0.87

agin

-0.86

Sag

-0.85

actor

-0.76

Johnson

-0.75

-0.74

API

-0.74

bag

-0.73

 federation

-0.73

POSITIVE LOGITS

Mon

1.79

Mon

1.73

MON

1.62

mon

1.43

MON

1.37

 Monte

1.33

Tue

1.24

mon

1.18

 Monarch

1.16

Tue

1.12

Activations Density 0.344%

words associated with actions or events happening in the past

punctuation and specific characters in text

No Comments

No Known Activations