INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

rand

-0.75

adj

-0.66

interrupted

-0.66

 interrupt

-0.64

riot

-0.63

 periodic

-0.63

zbek

-0.62

 livest

-0.62

nex

-0.62

¹

-0.62

POSITIVE LOGITS

owship

0.97

cellence

0.69

Appearance

0.66

alties

0.66

MpServer

0.66

stocks

0.65

ships

0.64

 Jackets

0.64

 Viking

0.62

boats

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.