INDEX

Explanations

numerical measurements and technical terms

oai_token-act-pair · gpt-3.5-turbo

terms related to events and emergencies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Cosponsors

-0.69

inventoryQuantity

-0.64

})

-0.60

Ĥİ

-0.59

jay

-0.58

uberty

-0.57

ort

-0.57

çĶŁ

-0.55

eals

-0.55

vid

-0.54

POSITIVE LOGITS

 lurking

0.79

 ready

0.75

 apiece

0.75

 leftover

0.73

 waiting

0.73

 amongst

0.72

 dotted

0.69

 besides

0.68

 wandering

0.68

 sighting

0.67

Activations Density 0.908%

numerical measurements and technical terms

terms related to events and emergencies

No Comments

No Known Activations