INDEX

Explanations

names of individuals with high prominence in various fields

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

proper nouns, specifically names of individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 lengths

-0.66

 understatement

-0.65

AAP

-0.64

 proving

-0.60

 anniversary

-0.60

 closure

-0.59

 NEXT

-0.58

 constant

-0.58

 abbrevi

-0.58

 confinement

-0.57

POSITIVE LOGITS

uman

1.26

plin

1.25

umann

1.25

efer

1.21

verty

1.21

urer

1.20

ube

1.16

udeau

1.15

ussen

1.14

essen

1.14

Activations Density 0.068%

names of individuals with high prominence in various fields

proper nouns, specifically names of individuals

No Comments

No Known Activations