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

 Bane

-0.84

interstitial

-0.83

 Chaser

-0.81

ombat

-0.80

 Sunder

-0.79

icter

-0.78

hens

-0.75

inho

-0.75

ãĥķãĤ©

-0.74

piracy

-0.74

POSITIVE LOGITS

 refin

0.70

 marriage

0.69

REAM

0.69

 omission

0.68

 commitment

0.67

 scholarship

0.67

 preference

0.66

 connection

0.66

 inacc

0.66

 society

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.