INDEX

Explanations

phrases related to capabilities or possibilities

oai_token-act-pair · gpt-3.5-turbo

phrases indicating quantities or amounts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 appointing

-0.69

 trustees

-0.69

 Hague

-0.68

 unveiling

-0.67

 Attribution

-0.64

 mediation

-0.64

ItemThumbnailImage

-0.64

 Tripoli

-0.64

 architects

-0.63

 attribution

-0.62

POSITIVE LOGITS

 crap

0.86

 batches

0.83

 stuff

0.82

stuff

0.81

ones

0.78

 kinds

0.75

aneously

0.75

aband

0.75

 consecut

0.74

onto

0.73

Activations Density 0.455%

phrases related to capabilities or possibilities

phrases indicating quantities or amounts

No Comments

No Known Activations