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

 Alonso

-0.67

 defect

-0.65

 Levant

-0.65

 separat

-0.64

 Isle

-0.63

usp

-0.62

 Republic

-0.62

 separatist

-0.62

 delet

-0.62

 curv

-0.61

POSITIVE LOGITS

swer

0.89

phia

0.76

amines

0.75

 ensued

0.74

atten

0.74

 Rollins

0.73

erd

0.72

ptives

0.70

ibilities

0.70

iculty

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.