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

è¦ļéĨĴ

-0.77

urring

-0.69

ihilation

-0.69

omics

-0.68

tein

-0.67

entially

-0.65

 Coat

-0.63

Bam

-0.63

anism

-0.63

 Balloon

-0.63

POSITIVE LOGITS

 shortages

0.69

 ensu

0.68

dq

0.68

 boil

0.68

 transsexual

0.67

 necessities

0.66

Haf

0.64

 glim

0.63

 spoiled

0.61

LIST

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.