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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Expedition

-0.73

yson

-0.71

awa

-0.65

oeuv

-0.64

uated

-0.64

uating

-0.62

 weaving

-0.62

Jac

-0.61

itton

-0.59

clips

-0.59

POSITIVE LOGITS

nir

0.76

theless

0.71

pora

0.70

pred

0.68

MAS

0.67

 scares

0.66

bath

0.66

ndum

0.65

SAN

0.64

dos

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.