INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

itia

-0.73

rollers

-0.68

 Compatibility

-0.62

sie

-0.62

 volleyball

-0.61

eem

-0.60

 eval

-0.59

poses

-0.59

aura

-0.57

zb

-0.57

POSITIVE LOGITS

ansk

0.87

ratulations

0.74

alach

0.73

ornings

0.69

odan

0.66

asar

0.65

inders

0.64

ldom

0.64

ovych

0.64

ãģ®å®

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.