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

ndra

-0.92

amacare

-0.79

ĵĺ

-0.78

ftime

-0.73

osher

-0.69

ħ

-0.69

office

-0.69

 drip

-0.68

wcs

-0.67

amily

-0.67

POSITIVE LOGITS

 Torrent

0.70

hog

0.66

Flor

0.65

 Nanto

0.65

Wem

0.64

Py

0.63

enge

0.62

elin

0.62

pat

0.62

ELD

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.