INDEX

Explanations

nouns related to various objects and activities, sometimes including specific nomenclature

oai_token-act-pair · gpt-3.5-turbo

phrases related to cooking and preparation

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

 confir

-0.74

£ı

-0.70

awa

-0.68

 stru

-0.67

 dismant

-0.63

 fourteen

-0.63

 furnish

-0.63

 reconc

-0.62

 exha

-0.62

 scaff

-0.61

POSITIVE LOGITS

1.60

1.40

Â²

1.09

1.00

II

0.99

second

0.93

0.91

II

0.89

0.88

Activations Density 0.129%

nouns related to various objects and activities, sometimes including specific nomenclature

phrases related to cooking and preparation

No Comments

No Known Activations