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

icter

-0.84

alach

-0.82

iae

-0.72

estead

-0.71

iability

-0.70

 sugg

-0.67

ikuman

-0.66

eki

-0.66

ahon

-0.66

 experien

-0.64

POSITIVE LOGITS

XL

0.74

.--

0.70

eln

0.69

 Push

0.68

Charges

0.66

lift

0.66

TM

0.64

 mega

0.63

 Version

0.62

 crore

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.