INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ESA

-0.78

stall

-0.75

aida

-0.74

 Emmy

-0.71

anye

-0.65

NCT

-0.64

 rounding

-0.62

 foreclosure

-0.62

dinand

-0.62

wall

-0.61

POSITIVE LOGITS

otos

0.77

zers

0.72

ipt

0.71

 McGu

0.68

 dare

0.68

 somet

0.68

uga

0.67

 furthermore

0.67

ouble

0.67

 Seconds

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.