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

 Label

-0.72

¥µ

-0.70

 Disclosure

-0.69

Pack

-0.69

 Leilan

-0.66

 Falcons

-0.66

 Invest

-0.65

 Ensure

-0.65

Begin

-0.64

>>>>

-0.64

POSITIVE LOGITS

 chromos

0.66

 bruises

0.63

TON

0.62

orah

0.61

 granddaughter

0.60

olls

0.60

sts

0.60

ija

0.59

ysc

0.59

ocaust

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.