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

flower

-0.82

ength

-0.74

velt

-0.74

ilts

-0.74

EStream

-0.68

renheit

-0.68

onew

-0.67

sburgh

-0.66

zona

-0.65

ibrary

-0.65

POSITIVE LOGITS

 Beir

0.71

 Bastard

0.62

icing

0.62

Ori

0.62

 rall

0.61

 minded

0.60

row

0.60

 unsu

0.59

 Herz

0.58

 wise

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.