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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Subjects

-0.82

 Instruct

-0.80

ebted

-0.72

Fig

-0.70

 constructs

-0.70

needs

-0.68

 Instit

-0.68

 Stats

-0.68

 Participants

-0.68

 Stru

-0.68

POSITIVE LOGITS

 nomine

0.76

VAT

0.70

 solicitor

0.68

pell

0.67

etheless

0.67

Ĥª

0.66

amacare

0.65

ulin

0.65

pour

0.64

 purge

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.