INDEX

Explanations

organizations or companies related to management or administration

oai_token-act-pair · gpt-3.5-turbo

terms related to management and organizations

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

 candle

-0.73

 stranger

-0.70

hi

-0.68

 canon

-0.67

 cruc

-0.67

 dirt

-0.66

 hero

-0.66

 coffin

-0.66

 extrad

-0.65

 playable

-0.64

POSITIVE LOGITS

 Management

3.68

Management

2.53

 management

2.27

management

2.13

 Manager

1.63

 manag

1.50

 Marketing

1.43

 Administration

1.38

 managers

1.35

 Managing

1.31

Activations Density 0.010%

organizations or companies related to management or administration

terms related to management and organizations

No Comments

No Known Activations