INDEX

Explanations

mentions of specific names like "Simpson" and "Scarborough."

oai_token-act-pair · gpt-3.5-turbo

proper nouns specifically names of individuals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Ware

-0.87

FC

-0.86

MW

-0.81

Rank

-0.80

reet

-0.80

GW

-0.72

eve

-0.72

 Gust

-0.72

 Warp

-0.70

 festival

-0.70

POSITIVE LOGITS

 Scarborough

1.83

 Simpson

1.66

 skeletons

1.59

 Lieberman

1.58

 fluor

1.49

 skeleton

1.47

 Norris

1.34

 Labrador

1.34

 Shirley

1.34

 fluoride

1.33

Activations Density 0.050%

mentions of specific names like "Simpson" and "Scarborough."

proper nouns specifically names of individuals

No Comments

No Known Activations