INDEX

Explanations

references to specific geographical locations, specifically Monaco and Luxembourg

oai_token-act-pair · gpt-3.5-turbo

mentions of specific geographic locations, particularly Monaco and Luxembourg

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

Ved

-0.75

 Tent

-0.69

 awaken

-0.68

 faculty

-0.64

Liter

-0.64

Viol

-0.63

 elic

-0.63

Wood

-0.63

 Eater

-0.63

 woke

-0.62

POSITIVE LOGITS

 Monaco

4.13

 Luxembourg

1.96

 Malta

1.90

 Cannes

1.79

 Bermuda

1.68

 Gibraltar

1.60

 Luxem

1.55

 Morocco

1.48

 Madagascar

1.41

 Bahamas

1.38

Activations Density 0.041%

references to specific geographical locations, specifically Monaco and Luxembourg

mentions of specific geographic locations, particularly Monaco and Luxembourg

No Comments

No Known Activations