INDEX

Explanations

keywords related to searching, urgency, information, and data analysis

oai_token-act-pair · gpt-3.5-turbo

words related to actions, responsibilities, and consequences in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

mma

-0.82

rique

-0.78

arks

-0.75

ELD

-0.70

oland

-0.69

 Tasman

-0.68

orthodox

-0.68

amen

-0.66

pedia

-0.65

ouched

-0.65

POSITIVE LOGITS

get

1.66

get

1.57

Get

1.49

Get

1.39

 gotten

1.38

GET

1.37

 gets

1.36

got

1.36

GET

1.30

 Gets

1.30

Activations Density 0.189%

keywords related to searching, urgency, information, and data analysis

words related to actions, responsibilities, and consequences in various contexts

No Comments

No Known Activations