INDEX

Explanations

noun phrases describing concepts, plans, or information

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

itely

-0.63

ortunately

-0.60

ternity

-0.58

die

-0.57

perty

-0.57

vertisement

-0.57

nesday

-0.57

ueller

-0.56

interrupted

-0.55

inia

-0.54

POSITIVE LOGITS

how

1.39

why

1.35

why

1.20

WHY

1.18

HOW

1.18

 whats

1.14

how

1.13

 what

1.09

what

1.03

How

0.95

Activations Density 0.096%

noun phrases describing concepts, plans, or information

No Comments

No Known Activations