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

ugal

-0.85

bid

-0.77

ighth

-0.76

odon

-0.68

obo

-0.68

 listings

-0.68

icter

-0.67

zn

-0.66

gnu

-0.65

bit

-0.65

POSITIVE LOGITS

·

0.81

ģ

0.71

rawdownloadcloneembedreportprint

0.68

estamp

0.68

 dressing

0.67

 accurately

0.65

 mold

0.63

 Dungeons

0.62

ĺ

0.60

VERTIS

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.