INDEX

Explanations

references to physical objects being grabbed

oai_token-act-pair · gpt-3.5-turbo

references to possession or objects associated with individuals

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

seek

-0.78

gallery

-0.69

eki

-0.67

umbered

-0.66

arter

-0.64

 omission

-0.64

eless

-0.62

uploads

-0.62

²¾

-0.62

clusive

-0.61

POSITIVE LOGITS

ById

0.92

 souven

0.77

hold

0.77

 likeness

0.70

 coveted

0.70

 priceless

0.68

 foothold

0.68

 trophies

0.67

ãĥł

0.67

 minerals

0.67

Activations Density 0.373%

references to physical objects being grabbed

references to possession or objects associated with individuals

No Comments

No Known Activations