INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ulf

-0.78

maximum

-0.71

COP

-0.71

 scoreboard

-0.70

 bragging

-0.68

yon

-0.64

yt

-0.63

 Pulitzer

-0.62

 HIGH

-0.62

 Oval

-0.61

POSITIVE LOGITS

Det

0.84

 regulars

0.75

ĸļ

0.73

abases

0.71

Kit

0.69

 compr

0.67

 Devi

0.67

 faiths

0.64

 ancest

0.64

ccording

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.