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

ufact

-0.84

senal

-0.73

 GOODMAN

-0.73

acebook

-0.72

 Pengu

-0.69

anmar

-0.69

inki

-0.68

TPPStreamerBot

-0.67

 Abortion

-0.66

theless

-0.65

POSITIVE LOGITS

obin

0.70

Naz

0.67

marked

0.63

 Than

0.61

tag

0.60

ateral

0.60

 snippet

0.60

saf

0.58

-(

0.58

ged

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.