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

bucks

-0.77

CDC

-0.68

Characters

-0.67

USD

-0.65

vote

-0.63

 Freem

-0.63

Vote

-0.63

rail

-0.63

elected

-0.62

 Toggle

-0.62

POSITIVE LOGITS

 Santos

0.74

ãĤ´ãĥ³

0.74

 Ambro

0.72

 seiz

0.71

çĲ

0.69

ãĥł

0.65

 Brune

0.65

ppo

0.65

 Beir

0.64

Roz

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.