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

roth

-0.76

bl

-0.70

 Leaves

-0.69

rays

-0.68

patient

-0.65

lying

-0.65

attled

-0.62

mosp

-0.61

mad

-0.61

antry

-0.61

POSITIVE LOGITS

ĸļ

0.68

nas

0.67

chwitz

0.67

STD

0.67

ovych

0.64

Reloaded

0.63

 Nemesis

0.63

Yi

0.62

ãĥĢ

0.61

 Palest

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.