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

 Inquisition

-0.68

 predator

-0.65

 birth

-0.61

 daddy

-0.61

 conception

-0.61

 organizer

-0.60

[/

-0.60

entity

-0.59

 Leilan

-0.58

 Demon

-0.58

POSITIVE LOGITS

acers

0.76

pees

0.75

BILITIES

0.72

amer

0.71

isites

0.69

adel

0.67

mic

0.67

nces

0.66

ENG

0.66

ufact

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.