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

 horizont

-0.76

eria

-0.69

 corrid

-0.67

tered

-0.65

UGE

-0.65

 Seym

-0.64

anos

-0.63

ornings

-0.63

 Lauder

-0.62

Mich

-0.61

POSITIVE LOGITS

 Permanent

0.75

phase

0.66

ellar

0.62

 Cognitive

0.62

rentice

0.62

body

0.61

 Election

0.61

 Arche

0.60

atche

0.59

 Stim

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.