INDEX

Explanations

the word "journey"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "ourn" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Vegeta

-0.70

 verse

-0.67

 Turbo

-0.64

 states

-0.63

 differentiation

-0.62

Boo

-0.62

 category

-0.59

 Mexican

-0.59

 contract

-0.58

 Moody

-0.58

POSITIVE LOGITS

ourn

4.73

ourning

2.12

ours

1.33

OUR

1.28

our

1.13

ourney

1.07

orde

1.06

acent

0.98

ouring

0.98

oun

0.97

Activations Density 0.012%

the word "journey"

occurrences of the word "ourn" and its variations

No Comments

No Known Activations