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

ioxide

-0.79

Rew

-0.70

Merit

-0.70

Ö¼

-0.70

Non

-0.66

Ni

-0.65

etc

-0.65

ģĸ

-0.65

enforcement

-0.65

ãĤ´ãĥ³

-0.64

POSITIVE LOGITS

 sights

0.74

PsyNetMessage

0.72

 prospects

0.65

 blast

0.65

 sparks

0.65

 session

0.64

ummies

0.64

 Ferr

0.63

 infiltr

0.63

 introdu

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.