INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

MpServer

-0.94

ipers

-0.93

atchewan

-0.81

antha

-0.80

encia

-0.79

BuyableInstoreAndOnline

-0.79

ulia

-0.77

igraph

-0.77

adelphia

-0.76

edin

-0.75

POSITIVE LOGITS

 core

0.91

 Core

0.79

ICAN

0.66

cru

0.66

 treatments

0.62

 chem

0.62

 containment

0.62

 trials

0.62

 clim

0.61

Īè

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.