INDEX

Explanations

references to the word "pol" with different suffixes

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ACTED

-0.68

INESS

-0.67

rament

-0.64

LY

-0.64

rers

-0.64

uring

-0.61

 Apprentice

-0.61

ORGE

-0.61

 Revel

-0.60

------------

-0.59

POSITIVE LOGITS

anski

1.19

itely

1.16

ikarp

1.02

ander

1.00

anco

1.00

iov

1.00

tical

0.99

arity

0.95

atility

0.95

ski

0.94

Activations Density 0.038%

references to the word "pol" with different suffixes

No Comments

No Known Activations