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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

coh

-0.76

 enrol

-0.71

 imperson

-0.71

 inform

-0.69

 expanding

-0.68

 reception

-0.67

 buzzing

-0.67

 picking

-0.66

 updating

-0.66

 stocking

-0.66

POSITIVE LOGITS

adr

0.93

strate

0.86

UGE

0.83

STEM

0.79

ares

0.79

ktop

0.77

inventoryQuantity

0.77

CU

0.77

uge

0.76

ublic

0.75

Activations Density 0.000%

No Known Activations

This feature has no known activations.