INDEX

Explanations

mentions of the "Supreme Leader Ayatollah Ali Khamenei."

oai_token-act-pair · gpt-3.5-turbo

references to Ayatollah Khamenei and his titles or attributes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Angels

-0.69

fox

-0.68

Ox

-0.66

 Oregon

-0.66

 Invisible

-0.65

 Virgin

-0.65

 Unreal

-0.63

 Edge

-0.61

 Missing

-0.59

 Sussex

-0.59

POSITIVE LOGITS

ollah

1.69

enei

1.12

 Seym

0.98

itory

0.85

 Kham

0.85

ength

0.83

ategory

0.83

aji

0.83

 slic

0.80

 lapt

0.79

Activations Density 0.009%

mentions of the "Supreme Leader Ayatollah Ali Khamenei."

references to Ayatollah Khamenei and his titles or attributes

No Comments

No Known Activations