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

etheless

-0.68

 distilled

-0.63

 Rodrig

-0.63

Mub

-0.62

 Bourbon

-0.62

 Martinez

-0.62

 notions

-0.62

 retrieval

-0.60

 hemor

-0.59

LR

-0.59

POSITIVE LOGITS

ing

0.77

ken

0.74

aber

0.67

iddling

0.66

edIn

0.66

obin

0.65

vr

0.65

edin

0.65

noon

0.65

ricted

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.