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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gins

-0.81

resy

-0.74

awk

-0.71

rolet

-0.71

lio

-0.69

uko

-0.68

nard

-0.66

Ĭ±

-0.63

wx

-0.62

illac

-0.62

POSITIVE LOGITS

anwhile

0.70

arching

0.64

 constit

0.61

aimon

0.60

ensation

0.59

eries

0.59

interstitial

0.59

theless

0.58

ANA

0.57

 privile

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.