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

ctory

-0.74

BLIC

-0.73

ramid

-0.73

ĸļ

-0.72

fred

-0.70

gently

-0.70

Reviewed

-0.68

tein

-0.68

tainment

-0.68

ternity

-0.67

POSITIVE LOGITS

 horsepower

0.71

ign

0.69

ulic

0.66

 traps

0.66

 vibrations

0.65

 Chak

0.64

ashtra

0.63

 Humane

0.62

mount

0.61

 Cunning

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.