INDEX

Explanations

references to the number of appearances or occurrences of something

oai_token-act-pair · gpt-3.5-turbo

references to the number of appearances 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

TYPE

-0.71

ress

-0.67

RA

-0.66

REF

-0.64

 Canal

-0.64

ARD

-0.64

 Peninsula

-0.63

IF

-0.63

RET

-0.62

Rocket

-0.62

POSITIVE LOGITS

uggest

1.51

ettings

1.23

hips

1.17

hip

1.11

poons

1.08

ourcing

1.07

paces

1.05

criptions

1.02

uits

1.02

etting

1.01

Activations Density 0.061%

references to the number of appearances or occurrences of something

references to the number of appearances in various contexts

No Comments

No Known Activations