INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

INESS

-0.93

fund

-0.79

ahon

-0.74

metics

-0.74

owitz

-0.71

ÂłÂłÂł

-0.70

MENTS

-0.68

SPONSORED

-0.68

regate

-0.66

avier

-0.65

POSITIVE LOGITS

 dwarves

0.75

 furnace

0.74

 timet

0.74

 torch

0.65

 Puppet

0.65

 Locked

0.64

 eclipse

0.63

 turtle

0.63

 wolf

0.63

 maid

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.