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

Iv

-0.76

Sher

-0.70

 Mahar

-0.67

sonian

-0.67

 Wasserman

-0.66

Zur

-0.65

Rah

-0.65

Pad

-0.65

Lur

-0.64

Pal

-0.63

POSITIVE LOGITS

cross

0.75

 guiActiveUnfocused

0.73

runs

0.72

 Gleaming

0.71

examination

0.70

devices

0.69

eligible

0.67

uploads

0.67

aying

0.66

adequ

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.