INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

edient

-0.65

 Polit

-0.65

©¶æ

-0.64

rete

-0.62

 Header

-0.62

aby

-0.60

 Bened

-0.60

 polled

-0.59

ãģł

-0.59

ewitness

-0.58

POSITIVE LOGITS

brand

0.67

potion

0.67

xon

0.65

 dwarves

0.64

TIT

0.62

ingen

0.61

castle

0.61

ATOR

0.61

Euro

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.