INDEX

Explanations

phrases related to personal or individual choices and inclinations

oai_token-act-pair · gpt-3.5-turbo

references to individual or collective preferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

bane

-0.75

kj

-0.73

Breaking

-0.71

CONCLUS

-0.69

angers

-0.68

amaz

-0.68

wordpress

-0.67

 Continental

-0.67

ãĥĥãĥī

-0.67

bleacher

-0.67

POSITIVE LOGITS

 preferences

1.08

yip

0.97

 preference

0.95

 favoring

0.88

 elig

0.81

 favoured

0.77

eering

0.76

 palate

0.75

 selection

0.73

 pane

0.73

Activations Density 0.012%

phrases related to personal or individual choices and inclinations

references to individual or collective preferences

No Comments

No Known Activations