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

ï¸

-0.76

 ðŁĳ

-0.72

arna

-0.72

 Machina

-0.71

orget

-0.69

sburgh

-0.68

oyle

-0.68

YPG

-0.67

eteria

-0.66

ieri

-0.66

POSITIVE LOGITS

EStreamFrame

0.94

dict

0.90

Abstract

0.81

dollar

0.78

Prime

0.77

LOCK

0.76

execute

0.75

bt

0.74

thro

0.73

constitutional

0.72

Activations Density 0.000%

No Known Activations

This feature has no known activations.