INDEX

Explanations

specific locations and geographic details

oai_token-act-pair · gpt-3.5-turbo

references to cities and geographical locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ads

-0.78

AUD

-0.77

andowski

-0.77

bugs

-0.73

 Parties

-0.73

Attack

-0.73

Episode

-0.72

intent

-0.71

Americans

-0.71

Instruct

-0.70

POSITIVE LOGITS

 mountainous

1.30

 suburb

1.27

 sprawling

1.25

 densely

1.22

 peninsula

1.18

 lush

1.18

 desolate

1.17

 swath

1.09

 bustling

1.09

 stronghold

1.08

Activations Density 0.222%

specific locations and geographic details

references to cities and geographical locations

No Comments

No Known Activations