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

jong

-0.75

========

-0.75

mor

-0.72

++++++++

-0.70

bur

-0.69

****************

-0.68

ulia

-0.67

++++++++++++++++

-0.67

glers

-0.66

 Gork

-0.65

POSITIVE LOGITS

ardless

0.72

oder

0.67

 Paint

0.67

osterone

0.66

Merit

0.66

ylon

0.65

ILE

0.65

ABV

0.64

BALL

0.64

DOM

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.