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

Bound

-0.77

Naz

-0.74

Box

-0.73

HOME

-0.69

Users

-0.68

Vote

-0.68

Control

-0.67

UP

-0.67

rius

-0.67

é¾įå

-0.66

POSITIVE LOGITS

 wonders

0.82

 fingerprints

0.68

 anomaly

0.67

abwe

0.66

 bead

0.65

 neighb

0.64

mble

0.63

utz

0.62

 Jagu

0.61

 assemb

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.