INDEX

Explanations

references to the CIA

oai_token-act-pair · gpt-3.5-turbo

references to the CIA

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

inances

-0.78

 Mulcair

-0.74

 Klopp

-0.70

alist

-0.70

 QuÃ©

-0.69

 Leafs

-0.68

à¼

-0.68

ï¸

-0.68

own

-0.67

 Indie

-0.67

POSITIVE LOGITS

 annex

0.87

 operatives

0.80

CIA

0.79

 Director

0.78

 covert

0.77

 operative

0.77

CIA

0.75

 assassinate

0.73

HQ

0.70

MS

0.70

Activations Density 0.011%

references to the CIA

references to the CIA

No Comments

No Known Activations