INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ãĥ¼ãĤ¯

-0.80

phia

-0.74

Ħ¢

-0.72

ãģ®éŃĶ

-0.72

comings

-0.68

DAQ

-0.67

orld

-0.66

Dia

-0.66

 acquisitions

-0.65

 merry

-0.65

POSITIVE LOGITS

 Brist

0.71

outh

0.68

 Scholars

0.68

Professor

0.67

otrop

0.65

ostics

0.63

 urgently

0.63

angs

0.59

thia

0.58

mbuds

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.