INDEX

Explanations

mentions of explosives or explosive-related terms

oai_token-act-pair · gpt-3.5-turbo

references to explosives and detonations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

aird

-0.78

Sea

-0.76

SEA

-0.75

 Vale

-0.74

 Seah

-0.72

TB

-0.72

 Ange

-0.71

 Dane

-0.70

Beast

-0.69

)=(

-0.67

POSITIVE LOGITS

 detonated

1.41

 deton

1.33

 explosives

1.18

 barric

1.04

 grenades

0.99

ilitary

0.98

 disposal

0.95

 deterrent

0.92

 bomb

0.92

 bomber

0.91

Activations Density 0.018%

mentions of explosives or explosive-related terms

references to explosives and detonations

No Comments

No Known Activations