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

 tremend

-0.80

 Dise

-0.80

 Cyan

-0.67

 Burning

-0.67

SOURCE

-0.65

Pv

-0.64

 Corpse

-0.64

lees

-0.64

 hous

-0.64

ulhu

-0.63

POSITIVE LOGITS

rang

0.80

Els

0.78

guided

0.71

gradient

0.71

sung

0.71

accept

0.69

dar

0.67

widget

0.67

GS

0.67

gart

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.