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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Specifications

-0.71

proven

-0.71

 Conv

-0.65

paying

-0.64

 Ratings

-0.64

 Instruction

-0.62

 Penet

-0.62

iciency

-0.61

 acquitted

-0.60

 appra

-0.60

POSITIVE LOGITS

ocument

0.76

 toast

0.71

Miller

0.70

emonium

0.70

Hunt

0.70

obyl

0.69

 Kislyak

0.68

PIN

0.67

orage

0.67

 Dmitry

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.