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

rendered

-0.69

-0.67

 Tucson

-0.66

RX

-0.65

obar

-0.64

itary

-0.63

 epoch

-0.63

ATI

-0.59

skinned

-0.58

 Mesa

-0.57

POSITIVE LOGITS

 Bagg

0.75

ulk

0.74

GREEN

0.73

mith

0.71

sbm

0.70

iane

0.69

bourg

0.66

ARR

0.65

gio

0.63

eele

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.