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

interstitial

-0.85

ivals

-0.83

omal

-0.79

comed

-0.78

 =================================

-0.77

acci

-0.77

airs

-0.73

ATIVE

-0.70

iblings

-0.70

 Grac

-0.68

POSITIVE LOGITS

 Jinn

0.68

Aad

0.66

ape

0.61

 wizard

0.60

SpaceEngineers

0.59

Xan

0.59

 powder

0.58

zzle

0.58

 Cobra

0.58

 Cyrus

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.