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

 Ancients

-0.75

éĹĺ

-0.72

çĶ°

-0.71

 plateau

-0.71

senal

-0.67

gorithm

-0.64

 proposition

-0.63

åĤ

-0.62

 travelling

-0.61

bral

-0.60

POSITIVE LOGITS

Ec

0.72

'"

0.68

 Bain

0.68

 Kamp

0.68

wik

0.68

 Gingrich

0.67

Corn

0.64

IPM

0.63

acc

0.63

Bud

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.