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

neys

-0.72

lost

-0.68

stic

-0.68

asts

-0.67

onis

-0.66

milo

-0.66

âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦âĢ¦

-0.65

 lashes

-0.64

ILE

-0.63

ripp

-0.63

POSITIVE LOGITS

 defic

0.75

 Warrant

0.66

 Lutheran

0.64

alities

0.63

 WARRANT

0.63

 waivers

0.63

 Amend

0.62

soDeliveryDate

0.62

acebook

0.59

ervative

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.