INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

rou

-0.74

ļéĨĴ

-0.70

odes

-0.69

grain

-0.69

OME

-0.69

orem

-0.68

VERTISEMENT

-0.68

 resil

-0.67

oros

-0.65

oha

-0.64

POSITIVE LOGITS

 albeit

0.98

 although

0.96

 however

0.93

 namely

0.87

 though

0.86

 including

0.86

 according

0.84

but

0.82

 which

0.82

 except

0.75

Activations Density 0.000%

No Known Activations

This feature has no known activations.