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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Enlightenment

-0.68

atures

-0.67

atives

-0.65

neys

-0.64

 Kelvin

-0.62

 plea

-0.61

 buds

-0.61

 wishes

-0.61

peak

-0.61

packages

-0.60

POSITIVE LOGITS

grave

0.73

aldo

0.71

torn

0.70

orman

0.67

icter

0.67

vironment

0.66

Gaming

0.66

OIL

0.66

achi

0.66

Ser

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.