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

ographies

-0.82

ibe

-0.71

 scrut

-0.69

jri

-0.69

oter

-0.68

channelAvailability

-0.66

ications

-0.65

ylan

-0.64

 defic

-0.64

oters

-0.64

POSITIVE LOGITS

nda

0.66

CU

0.66

 Mata

0.66

Times

0.64

 Frozen

0.63

ĳ

0.63

TED

0.62

cult

0.61

 Grab

0.61

ional

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.