INDEX

Explanations

proper nouns or names (e.g., Susie, Susana, Suzette) in a text

oai_token-act-pair · gpt-3.5-turbo

references to a specific character or name related to suspicion or scrutiny

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

è¦ļéĨĴ

-0.94

akings

-0.68

 Doomsday

-0.65

 sorting

-0.64

 dividing

-0.64

 overhead

-0.64

 separating

-0.63

 stakes

-0.63

 Kinnikuman

-0.63

 living

-0.62

POSITIVE LOGITS

pected

1.33

pect

1.30

pects

1.23

pecting

1.21

pic

1.21

annah

1.14

pir

1.14

cept

1.01

cription

0.91

sex

0.91

Activations Density 0.022%

proper nouns or names (e.g., Susie, Susana, Suzette) in a text

references to a specific character or name related to suspicion or scrutiny

No Comments

No Known Activations