INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

croft

-0.79

books

-0.69

sat

-0.66

successful

-0.66

balls

-0.65

cule

-0.65

law

-0.62

ãĤ¶

-0.61

renheit

-0.61

gebra

-0.61

POSITIVE LOGITS

Ń·

0.75

 VIDEOS

0.74

ILA

0.72

ï¸

0.69

EG

0.68

owship

0.68

 Unlimited

0.66

IME

0.65

IDES

0.65

INO

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.