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

ModLoader

-0.72

relevant

-0.69

orsi

-0.68

etheless

-0.66

etheus

-0.65

 maximizing

-0.63

ower

-0.61

adjusted

-0.61

HAEL

-0.60

izable

-0.60

POSITIVE LOGITS

ogly

0.86

atel

0.80

unct

0.78

 Cosponsors

0.71

è£ħ

0.70

wal

0.70

abet

0.69

stan

0.68

zens

0.67

lez

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.