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

MENTS

-0.68

 Emin

-0.67

ele

-0.67

eds

-0.66

guards

-0.66

 descriptions

-0.65

 prayers

-0.63

ãĥ«

-0.63

Guard

-0.62

ahime

-0.61

POSITIVE LOGITS

©¶æ

0.85

Bus

0.69

kefeller

0.67

Jaw

0.67

olitics

0.65

hattan

0.65

Ray

0.64

ublic

0.64

Pis

0.64

 Triangle

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.