INDEX

Explanations

phrases indicating causality or influence

oai_token-act-pair · gpt-3.5-turbo

phrases indicating cause-and-effect relationships

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

alty

-0.64

 Lethal

-0.63

 ledge

-0.62

 Nationals

-0.61

 Clarks

-0.61

rament

-0.60

 Bleach

-0.60

 Cedar

-0.58

 Essential

-0.57

 Challenger

-0.57

POSITIVE LOGITS

SPONSORED

0.83

________________________________________________________________

0.79

Pub

0.75

PsyNetMessage

0.72

FTWARE

0.72

ãĥĥãĥī

0.70

oros

0.68

Shape

0.68

nings

0.67

DragonMagazine

0.67

Activations Density 0.011%

phrases indicating causality or influence

phrases indicating cause-and-effect relationships

No Comments

No Known Activations