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

EStream

-0.71

forward

-0.71

ername

-0.70

adapt

-0.68

 compr

-0.67

rek

-0.67

icter

-0.67

ãĤ§

-0.66

quished

-0.65

hement

-0.65

POSITIVE LOGITS

 Haram

0.80

olson

0.66

 rarity

0.63

amba

0.62

uala

0.62

yon

0.61

hya

0.61

 Petr

0.61

RBI

0.61

 Territory

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.