INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ully

-0.71

dinand

-0.69

ynes

-0.64

ocalypse

-0.64

ocaly

-0.63

rys

-0.63

|--

-0.62

FTA

-0.62

ymph

-0.62

arent

-0.62

POSITIVE LOGITS

dar

0.69

ETA

0.65

å°Ĩ

0.65

 fixation

0.62

 barking

0.61

ó

0.61

Die

0.60

fix

0.60

 commenters

0.60

 buds

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.