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

chnology

-0.80

seekers

-0.71

gaard

-0.71

INESS

-0.71

Reward

-0.69

Magikarp

-0.68

FORMATION

-0.67

 Tsukuyomi

-0.67

phia

-0.66

worthiness

-0.66

POSITIVE LOGITS

 retweet

0.68

 butcher

0.67

ammy

0.65

obo

0.63

wk

0.63

 dumb

0.63

mar

0.63

 prep

0.62

artment

0.61

ĨĴ

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.