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

tal

-0.70

 Read

-0.67

âĹı

-0.64

Putin

-0.61

UGC

-0.61

inel

-0.61

âĪ

-0.61

ante

-0.60

SPONSORED

-0.60

allow

-0.59

POSITIVE LOGITS

¥ŀ

0.73

©¶æ

0.73

velt

0.72

 tremend

0.72

insula

0.71

 lifes

0.71

 sizing

0.68

 silhou

0.68

 diagram

0.67

ascript

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.