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

 cannons

-0.72

 poisons

-0.69

 poisoned

-0.68

 reprodu

-0.68

pse

-0.66

 shelling

-0.65

 polls

-0.63

 fertile

-0.63

 poll

-0.63

 Poll

-0.62

POSITIVE LOGITS

PsyNetMessage

1.04

MSN

0.77

chel

0.76

å¥³

0.75

gie

0.71

Constructed

0.70

pez

0.69

Dom

0.69

HUD

0.68

atars

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.