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

achus

-0.77

vertising

-0.75

Age

-0.68

âĹ¼

-0.66

********

-0.65

acts

-0.65

jing

-0.64

nea

-0.64

ipl

-0.63

boards

-0.63

POSITIVE LOGITS

 Homs

0.65

 narrated

0.62

iott

0.61

ciples

0.61

 Maduro

0.61

urally

0.61

Mk

0.60

huh

0.59

?).

0.56

 Purg

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.