INDEX

Explanations

mentions of the word "Des"

oai_token-act-pair · gpt-3.5-turbo

occurrence of the descriptor "Des" followed by various numbers or contexts related to descriptions or titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

glers

-0.85

Reviewer

-0.74

hetti

-0.64

 Kinnikuman

-0.63

heartedly

-0.63

 shake

-0.59

 overdue

-0.59

 altogether

-0.59

 Penguin

-0.58

 Klaus

-0.58

POSITIVE LOGITS

 Moines

1.36

cend

1.33

perate

1.26

erve

1.24

cribed

1.23

criptions

1.21

erves

1.20

irable

1.15

ignt

1.14

cript

1.08

Activations Density 0.017%

mentions of the word "Des"

occurrence of the descriptor "Des" followed by various numbers or contexts related to descriptions or titles

No Comments

No Known Activations