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

onsense

-0.72

ILA

-0.71

é¾į

-0.67

UFC

-0.63

EMBER

-0.62

ANGE

-0.62

 madness

-0.61

 Ukip

-0.60

Beast

-0.60

æĪ¦

-0.60

POSITIVE LOGITS

ources

0.71

 Cipher

0.69

 captcha

0.65

translation

0.65

 trave

0.64

vis

0.63

 surpr

0.62

travel

0.62

uctor

0.61

ufact

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.