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

 Ezra

-0.72

 boobs

-0.64

 scrimmage

-0.64

 microphones

-0.63

 Advertisement

-0.63

clips

-0.62

gren

-0.62

Ø±

-0.62

[+

-0.61

 beads

-0.61

POSITIVE LOGITS

atis

0.85

jriwal

0.76

alach

0.76

ain

0.75

erest

0.75

irtual

0.74

ModLoader

0.74

chy

0.73

anium

0.73

chwitz

0.73

Activations Density 0.000%

No Known Activations

This feature has no known activations.