INDEX

Explanations

instructions or descriptions related to computer programming, particularly focusing on input and output operations

oai_token-act-pair · gpt-3.5-turbo

phrases related to output specifications or results within a technical context

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

creen

-0.69

Fort

-0.63

tones

-0.63

hra

-0.61

hon

-0.61

 Rugby

-0.60

kers

-0.60

 Spit

-0.59

 Shore

-0.58

western

-0.58

POSITIVE LOGITS

 output

1.33

 outputs

1.30

 Output

1.13

output

1.07

ilib

0.94

rious

0.93

Output

0.92

flows

0.83

ted

0.80

alted

0.78

Activations Density 0.007%

instructions or descriptions related to computer programming, particularly focusing on input and output operations

phrases related to output specifications or results within a technical context

No Comments

No Known Activations