INDEX

Explanations

verbs related to causing a negative physical reaction or harm

oai_token-act-pair · gpt-3.5-turbo

words related to emphasis or intensity, particularly those that include the suffix 'inate.'

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Merit

-0.61

 Whit

-0.58

tug

-0.58

 safety

-0.56

 diver

-0.55

 attendant

-0.54

 Link

-0.53

 history

-0.53

Mer

-0.52

 offered

-0.52

POSITIVE LOGITS

inate

4.68

inates

3.14

inately

2.84

inating

2.83

inated

2.73

ination

2.36

inators

1.94

inations

1.91

inator

1.68

inatory

1.48

Activations Density 0.005%

verbs related to causing a negative physical reaction or harm

words related to emphasis or intensity, particularly those that include the suffix 'inate.'

No Comments

No Known Activations