INDEX

Explanations

personal pronouns referring to individuals

oai_token-act-pair · gpt-3.5-turbo

references to individuals in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

endif

-0.86

 expiration

-0.73

 antioxid

-0.65

 expire

-0.63

stay

-0.62

NG

-0.61

 disappear

-0.59

 endif

-0.59

ENSE

-0.59

noon

-0.59

POSITIVE LOGITS

 deems

1.18

aped

1.17

uristic

1.14

 considers

1.09

 oversaw

1.07

 blames

1.05

 describes

1.04

 believes

1.04

 accuses

0.99

 alleges

0.99

Activations Density 0.113%

personal pronouns referring to individuals

references to individuals in the text

No Comments

No Known Activations