INDEX

Explanations

mentions of specific locations or institutions, particularly universities, cities, and government offices

oai_token-act-pair · gpt-3.5-turbo

place names and locations

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

vice

-0.84

-0.81

lihood

-0.77

rave

-0.76

razen

-0.75

Ireland

-0.74

bre

-0.72

obyl

-0.72

epad

-0.72

 Nguyen

-0.72

POSITIVE LOGITS

 Independence

0.69

 overrun

0.67

 unpop

0.67

 districts

0.63

 Conserv

0.63

 segregated

0.63

 countryside

0.63

 Division

0.63

 Canal

0.63

 Council

0.62

Activations Density 0.348%

mentions of specific locations or institutions, particularly universities, cities, and government offices

place names and locations

No Comments

No Known Activations