INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Isis

-0.72

Maker

-0.67

existence

-0.67

arta

-0.66

 Metatron

-0.65

Alm

-0.65

 Israelis

-0.64

 Maker

-0.63

Elsa

-0.63

 Nieto

-0.62

POSITIVE LOGITS

ntil

0.92

redit

0.78

undai

0.78

rawdownloadcloneembedreportprint

0.77

IOR

0.70

 graded

0.69

 satur

0.68

 fined

0.68

etheless

0.68

hops

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.