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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Fal

-0.80

Ļ½

-0.71

MI

-0.64

mone

-0.64

ģ«

-0.63

©¶æ¥µ

-0.62

©¶æ

-0.61

ãģ®æ

-0.61

arc

-0.61

 Hitman

-0.61

POSITIVE LOGITS

lig

0.77

Â¨

0.72

ignt

0.67

tel

0.66

lands

0.64

lations

0.63

landers

0.63

 Christy

0.62

leans

0.61

essee

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.