INDEX

Explanations

No Explanations Found

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

anan

-0.87

 downs

-0.70

apixel

-0.63

ses

-0.62

 results

-0.59

NF

-0.59

 Guinness

-0.58

pse

-0.57

 record

-0.57

pees

-0.56

POSITIVE LOGITS

mt

0.78

agent

0.77

opol

0.67

mob

0.66

GU

0.65

Staff

0.64

quished

0.64

ctors

0.64

Lib

0.63

eligible

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.