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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 millenn

-0.71

 adoption

-0.66

NetMessage

-0.64

 advoc

-0.64

zik

-0.63

 poses

-0.62

 proponent

-0.62

zin

-0.61

 titan

-0.61

 dispens

-0.61

POSITIVE LOGITS

ensibly

0.78

mbuds

0.77

ptions

0.72

liam

0.69

ithmetic

0.69

estyles

0.67

ply

0.66

merce

0.65

appiness

0.65

 Average

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.