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

ean

-0.67

 paradise

-0.64

 Chill

-0.64

pace

-0.63

 cathedral

-0.61

 climb

-0.61

 Celestial

-0.60

CHAT

-0.59

 boredom

-0.59

 ascend

-0.58

POSITIVE LOGITS

pell

0.74

vernment

0.73

abouts

0.72

illion

0.69

cules

0.67

agall

0.67

bos

0.65

zsche

0.65

lez

0.64

agents

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.