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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Ratings

-0.63

©¶æ

-0.62

 Mane

-0.62

 annex

-0.61

rous

-0.60

 Ambro

-0.60

hover

-0.59

ffe

-0.58

arters

-0.58

adj

-0.57

POSITIVE LOGITS

ogene

0.81

ledge

0.74

vity

0.67

ãĥīãĥ©

0.66

orers

0.64

umbn

0.64

gets

0.63

ãĤ£

0.62

ESH

0.62

NO

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.