INDEX

Explanations

phrases related to cultural or societal norms

oai_token-act-pair · gpt-3.5-turbo

references to social norms and their variations

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

 Lama

-0.68

semble

-0.66

 Kush

-0.62

 Sunder

-0.59

 Newport

-0.58

wrapper

-0.58

 istg

-0.58

Tub

-0.57

 Lizard

-0.57

 Riverside

-0.57

POSITIVE LOGITS

ativity

1.31

ality

1.18

 ante

0.97

atively

0.92

als

0.85

quo

0.80

itionally

0.79

eers

0.79

 prev

0.76

essential

0.75

Activations Density 0.035%

phrases related to cultural or societal norms

references to social norms and their variations

No Comments

No Known Activations