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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Sind

-0.69

 retrospect

-0.66

 Sith

-0.66

Haf

-0.64

soDeliveryDate

-0.63

 Amon

-0.62

borg

-0.61

rite

-0.61

 hereafter

-0.60

gad

-0.60

POSITIVE LOGITS

uminati

0.78

ouring

0.75

urses

0.73

ggies

0.70

odynamics

0.69

ockets

0.68

agus

0.67

ograp

0.65

 Sovere

0.64

Queue

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.