INDEX

Explanations

high-ranking military titles

oai_token-act-pair · gpt-3.5-turbo

military ranks and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 tumble

-0.78

 fracture

-0.75

 pudding

-0.72

 crash

-0.69

 collide

-0.68

 panc

-0.63

Consumer

-0.63

 fractures

-0.62

 Pupp

-0.62

 cruc

-0.62

POSITIVE LOGITS

 Joseph

0.92

Lt

0.90

 Hasan

0.87

Maj

0.87

 Samuel

0.87

 John

0.85

Lt

0.85

 Edwin

0.85

 Jeffrey

0.84

essors

0.84

Activations Density 0.086%

high-ranking military titles

military ranks and titles

No Comments

No Known Activations