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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ohan

-0.72

ombat

-0.71

enna

-0.64

ussie

-0.64

erto

-0.63

 Conquer

-0.59

coh

-0.58

 Britann

-0.57

REE

-0.57

croft

-0.57

POSITIVE LOGITS

¶ħ

0.64

ocide

0.63

iphate

0.62

sha

0.62

 spons

0.62

exempt

0.61

Mir

0.58

kees

0.58

shit

0.57

pmwiki

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.