INDEX

Explanations

words related to organizational or business management

oai_token-act-pair · gpt-3.5-turbo

references to management-related concepts and practices

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

arthed

-0.72

ipples

-0.67

anon

-0.65

oaded

-0.64

Joined

-0.63

itate

-0.63

eto

-0.63

III

-0.62

 Innocent

-0.62

eps

-0.61

POSITIVE LOGITS

 management

3.68

management

2.90

 Management

2.81

Management

2.79

 manag

2.50

 managers

2.25

 manager

1.83

manager

1.77

 managing

1.62

 managerial

1.57

Activations Density 0.022%

words related to organizational or business management

references to management-related concepts and practices

No Comments

No Known Activations