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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 recess

-0.75

maj

-0.72

olson

-0.71

lux

-0.71

 subordinates

-0.70

 advoc

-0.67

gettable

-0.66

 disgu

-0.65

 elim

-0.64

 superiors

-0.63

POSITIVE LOGITS

MSM

0.77

 Processing

0.73

iru

0.71

 Ginger

0.68

 Pigs

0.67

ĺħ

0.65

 Tanz

0.65

 Farming

0.64

 Plymouth

0.64

ahu

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.