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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 inclusive

-0.69

oldemort

-0.68

ois

-0.67

 viol

-0.66

Ü

-0.65

onyms

-0.64

 actionGroup

-0.64

idi

-0.63

 Uran

-0.63

ackets

-0.62

POSITIVE LOGITS

patient

0.81

ove

0.70

TPPStreamerBot

0.69

BALL

0.67

hops

0.67

licks

0.67

glomer

0.66

biz

0.65

statement

0.64

club

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.