INDEX

Explanations

words related to the purpose or intent behind actions or entities

oai_token-act-pair · gpt-3.5-turbo

references to the concept of purpose

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Chance

-0.65

Pick

-0.65

 Leopard

-0.64

 Drops

-0.62

alus

-0.61

 Temper

-0.61

 slopes

-0.61

aq

-0.60

 Polo

-0.59

igans

-0.58

POSITIVE LOGITS

ful

1.50

fully

1.23

fulness

1.13

FUL

1.01

lessness

0.98

full

0.96

lessly

0.93

ories

0.83

²¾

0.80

less

0.77

Activations Density 0.064%

words related to the purpose or intent behind actions or entities

references to the concept of purpose

No Comments

No Known Activations