INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ĸļ

-0.93

hyde

-0.82

©¶æ¥µ

-0.72

length

-0.68

irit

-0.68

notation

-0.67

oÄŁ

-0.66

shed

-0.65

æĦ

-0.65

AAF

-0.65

POSITIVE LOGITS

mug

0.70

 Walmart

0.68

 Rhodes

0.67

 Euras

0.65

 Zimbabwe

0.64

 Venezuela

0.63

 Stamford

0.63

ITAL

0.62

 voucher

0.61

 Ukraine

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.