INDEX

Explanations

words related to the beginning or initiation of activities or processes

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the initiation of actions or processes

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

xual

-0.74

justice

-0.68

vich

-0.67

dain

-0.65

mun

-0.65

uro

-0.63

ylum

-0.63

alia

-0.61

forums

-0.60

beit

-0.60

POSITIVE LOGITS

 start

3.47

start

2.28

 Start

2.28

 begin

2.27

 starts

2.06

 beginning

1.94

Start

1.93

 starting

1.82

 START

1.79

 commence

1.70

Activations Density 0.024%

words related to the beginning or initiation of activities or processes

phrases indicating the initiation of actions or processes

No Comments

No Known Activations