INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

geist

-0.77

aughs

-0.75

coon

-0.74

ubi

-0.73

ortex

-0.72

ilts

-0.70

insula

-0.69

oola

-0.68

livious

-0.68

 tentacles

-0.67

POSITIVE LOGITS

\/

0.69

Unknown

0.63

 larg

0.62

vP

0.62

 scarcity

0.62

 attribution

0.61

 Merc

0.60

Loading

0.59

 foll

0.59

PK

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.