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

 Cycle

-0.69

Sv

-0.67

afety

-0.67

raph

-0.65

emaker

-0.65

erer

-0.64

makers

-0.64

INAL

-0.64

ethe

-0.63

CTR

-0.63

POSITIVE LOGITS

haz

0.70

 necks

0.67

 refriger

0.63

haul

0.62

 handwriting

0.62

sou

0.61

 roasted

0.61

Written

0.60

 departing

0.60

Leaks

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.