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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

CHAT

-0.65

 liquid

-0.63

 privat

-0.63

 redistributed

-0.62

tur

-0.61

 tattoo

-0.60

²¾

-0.59

 camps

-0.57

 parap

-0.57

 claw

-0.57

POSITIVE LOGITS

hma

0.87

ridge

0.84

hetti

0.83

rique

0.76

riad

0.76

fare

0.75

rou

0.75

gling

0.74

odore

0.73

azard

0.72

Activations Density 0.000%

No Known Activations

This feature has no known activations.