INDEX

Explanations

phrases related to packaging or packing

oai_token-act-pair · gpt-3.5-turbo

references to grouped items or collections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 adultery

-0.74

 Benef

-0.71

Byr

-0.69

 precedent

-0.68

Lia

-0.68

 Shia

-0.67

 Inqu

-0.66

NCT

-0.65

 ausp

-0.65

 resonance

-0.64

POSITIVE LOGITS

pack

4.38

packs

2.85

Pack

2.48

 pack

2.21

packing

2.05

 Pack

1.99

 packs

1.64

packed

1.63

 PACK

1.53

 Packs

1.46

Activations Density 0.009%

phrases related to packaging or packing

references to grouped items or collections

No Comments

No Known Activations