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

MON

-0.68

pub

-0.67

 Springer

-0.65

NS

-0.64

PAR

-0.62

TOR

-0.62

 Univ

-0.61

 Rosenthal

-0.61

Fra

-0.60

Tickets

-0.60

POSITIVE LOGITS

odox

0.87

iously

0.86

ahime

0.83

iple

0.81

etheless

0.77

nect

0.77

selage

0.77

ipher

0.77

uries

0.75

inge

0.75

Activations Density 0.000%

No Known Activations

This feature has no known activations.