INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Tanks

-0.69

 Tank

-0.68

 Disp

-0.68

 Paradise

-0.65

pav

-0.65

 Cathedral

-0.64

iche

-0.64

 Battalion

-0.63

 Hast

-0.62

idia

-0.62

POSITIVE LOGITS

regular

0.82

pees

0.76

LU

0.75

vin

0.72

virt

0.72

alias

0.70

rots

0.70

keyes

0.70

-------

0.69

lu

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.