INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

MVP

-0.71

DOI

-0.66

internal

-0.64

 rematch

-0.64

ierrez

-0.63

 amygdala

-0.62

 ILCS

-0.61

inged

-0.61

rage

-0.61

 Emails

-0.60

POSITIVE LOGITS

edom

0.90

livious

0.73

byn

0.71

kef

0.69

xual

0.67

bilt

0.64

wegian

0.63

 distingu

0.63

 guiActiveUn

0.63

reek

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.