INDEX

Explanations

references to diplomats and diplomatic activities

oai_token-act-pair · gpt-3.5-turbo

references to diplomats and diplomatic roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

alach

-0.73

 Bore

-0.66

forest

-0.66

 Alto

-0.65

 Progressive

-0.64

Enough

-0.64

 Hath

-0.63

efer

-0.63

vent

-0.62

 Passage

-0.62

POSITIVE LOGITS

 diplomats

1.02

 diplomat

1.02

 corps

0.95

glers

0.87

 diplomacy

0.84

 envoy

0.80

 diplomatic

0.80

 Lavrov

0.80

clinton

0.79

 abroad

0.78

Activations Density 0.013%

references to diplomats and diplomatic activities

references to diplomats and diplomatic roles

No Comments

No Known Activations