INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ricted

-0.72

uffs

-0.71

raining

-0.70

liga

-0.70

á¹

-0.69

uffle

-0.68

lux

-0.68

È

-0.67

 Strikes

-0.65

olars

-0.64

POSITIVE LOGITS

orah

0.63

 horizon

0.62

 monop

0.60

 granddaughter

0.58

OM

0.58

 GMOs

0.58

iewicz

0.58

onia

0.57

mary

0.57

 vide

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.