INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Rog

-0.74

ctic

-0.70

dden

-0.69

Esk

-0.65

 fortun

-0.65

scl

-0.64

external

-0.63

orean

-0.62

terday

-0.61

existent

-0.61

POSITIVE LOGITS

lees

0.78

 Disciple

0.69

 Indra

0.69

 Guru

0.68

 Label

0.68

 Berkshire

0.67

 Heisman

0.67

ucket

0.66

 Beans

0.66

 Tuls

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.