INDEX

Explanations

mentions of the word "employees."

oai_token-act-pair · gpt-3.5-turbo

references to employees in various contexts

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

mire

-0.67

MOD

-0.67

 Babel

-0.64

olon

-0.63

Picture

-0.63

 spoiler

-0.62

psy

-0.61

rav

-0.60

colour

-0.60

uri

-0.60

POSITIVE LOGITS

 employees

3.50

 Employees

2.81

 employee

2.55

 staffers

2.31

 workers

2.25

employ

2.12

 coworkers

1.96

 Employee

1.93

 executives

1.91

Employ

1.88

Activations Density 0.021%

mentions of the word "employees."

references to employees in various contexts

No Comments

No Known Activations