INDEX

Explanations

government-related terms and policies

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to governmental actions, policies, or regulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 latch

-0.79

 memory

-0.79

 Zeit

-0.75

 Lives

-0.73

 slopes

-0.72

 swim

-0.72

die

-0.71

 modesty

-0.71

 memories

-0.71

 inhabit

-0.71

POSITIVE LOGITS

approved

1.83

issued

1.78

imposed

1.77

sponsored

1.75

provided

1.74

appointed

1.70

assisted

1.66

backed

1.66

ordered

1.65

authorized

1.65

Activations Density 0.062%

government-related terms and policies

references to governmental actions, policies, or regulations

No Comments

No Known Activations