INDEX

Explanations

entities related to military and government positions or organizations

oai_token-act-pair · gpt-3.5-turbo

references to military ranks and positions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 stub

-0.74

rists

-0.66

 dazz

-0.65

 itch

-0.65

bub

-0.65

 lengths

-0.63

 piling

-0.62

 foresee

-0.62

 raining

-0.62

 setbacks

-0.62

POSITIVE LOGITS

Lia

0.73

ĸļ

0.70

 Operations

0.69

 Strategic

0.69

 Affairs

0.67

WA

0.67

iciary

0.67

 advisory

0.66

 Chiefs

0.66

ederation

0.65

Activations Density 0.508%

entities related to military and government positions or organizations

references to military ranks and positions

No Comments

No Known Activations