INDEX

Explanations

words related to specific geographical locations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating locations or proximity to various areas

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

undrum

-0.78

 partName

-0.72

Rust

-0.70

atis

-0.68

DAQ

-0.67

mat

-0.66

uten

-0.65

Trend

-0.65

Disk

-0.61

 Slay

-0.61

POSITIVE LOGITS

 alike

0.86

 respectively

0.63

 commencement

0.59

 industries

0.59

)=(

0.59

 ceremonies

0.58

 plantations

0.58

 deployments

0.57

soDeliveryDate

0.56

 emergencies

0.55

Activations Density 0.062%

words related to specific geographical locations

phrases indicating locations or proximity to various areas

No Comments

No Known Activations