INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

natureconservancy

-0.72

notations

-0.68

Gre

-0.68

cles

-0.67

 contributors

-0.67

ocene

-0.66

cov

-0.66

Benef

-0.66

 anchors

-0.65

 anchor

-0.64

POSITIVE LOGITS

ãĥ©ãĥ³

0.78

 Swordsman

0.73

anish

0.67

istani

0.66

ardless

0.64

reme

0.64

acket

0.63

rad

0.63

rew

0.61

 Puzz

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.