INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

è¦ļéĨĴ

-0.81

soDeliveryDate

-0.72

jar

-0.72

jury

-0.70

ãĥ¯ãĥ³

-0.70

--------------------

-0.69

 Archdemon

-0.67

ãĥ¼ãĥĨãĤ£

-0.66

IB

-0.66

POSITIVE LOGITS

GOD

0.66

 worthless

0.61

 abdom

0.60

 Dupl

0.58

gencies

0.57

 Emerson

0.57

oton

0.57

 Chrom

0.56

ussia

0.56

chn

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.