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

 delegation

-0.87

ancers

-0.78

rawdownloadcloneembedreportprint

-0.75

 Ancients

-0.71

 palette

-0.69

ixture

-0.66

berra

-0.66

gemony

-0.65

oshi

-0.64

pora

-0.64

POSITIVE LOGITS

 NETWORK

0.69

 Canad

0.65

 Wilson

0.62

fox

0.60

 Fern

0.60

 Cedar

0.59

 Beacon

0.59

 fragmented

0.58

 Judy

0.58

Dispatch

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.