INDEX

Explanations

proper nouns representing names of people

oai_token-act-pair · gpt-3.5-turbo

references to notable individuals and their professional roles or titles

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

Gaza

-0.84

ahime

-0.82

 selfie

-0.81

angan

-0.77

bia

-0.75

awatts

-0.75

izons

-0.75

 rollout

-0.74

inki

-0.73

 selfies

-0.73

POSITIVE LOGITS

 William

1.50

 Herbert

1.50

 Harold

1.48

 Ernest

1.42

 Sidney

1.41

 Fritz

1.41

 Charles

1.41

 Frederick

1.40

 Ralph

1.40

 Arthur

1.39

Activations Density 0.266%

proper nouns representing names of people

references to notable individuals and their professional roles or titles

No Comments

No Known Activations