INDEX

Explanations

phrases related to differentiating or grouping items or concepts

oai_token-act-pair · gpt-3.5-turbo

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

olph

-0.72

;}

-0.70

ilus

-0.64

pex

-0.64

opa

-0.62

ogi

-0.62

Wiz

-0.60

Äĩ

-0.60

Surv

-0.60

 Weld

-0.59

POSITIVE LOGITS

 circumstances

0.99

 contexts

0.97

 sizes

0.94

 manner

0.94

 climates

0.92

 circumstance

0.91

 styles

0.91

 sexes

0.87

hra

0.86

 eras

0.85

Activations Density 0.193%

phrases related to differentiating or grouping items or concepts

No Comments

No Known Activations