INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

xual

-0.75

involved

-0.70

 contrace

-0.70

ertodd

-0.67

cipled

-0.66

²¾

-0.65

 shooters

-0.63

birds

-0.63

 hinges

-0.61

 avert

-0.60

POSITIVE LOGITS

wana

0.81

 Gadget

0.80

 Sphere

0.74

 Cortex

0.66

ghan

0.64

 Runtime

0.64

ALLY

0.64

 Frontier

0.63

 Kraken

0.62

 distortion

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.