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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Wars

-0.84

Fed

-0.71

 Emin

-0.70

Wim

-0.68

 Reserv

-0.65

Cub

-0.65

Jenn

-0.65

Tid

-0.65

 infertility

-0.64

 Morton

-0.64

POSITIVE LOGITS

©¶æ

0.80

 Bulldogs

0.75

ername

0.72

EngineDebug

0.71

 indic

0.68

expression

0.68

eta

0.66

arest

0.66

maxwell

0.66

NCT

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.