INDEX

Explanations

military ranks and titles

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Lieutenant" and its abbreviation "Lt."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

gem

-0.69

 chop

-0.68

holding

-0.64

aves

-0.64

conom

-0.64

raft

-0.63

selves

-0.63

banks

-0.62

 continents

-0.60

Cra

-0.59

POSITIVE LOGITS

Lt

3.68

Lt

3.31

 Lieutenant

2.80

 lieutenant

2.50

Sgt

1.83

utenant

1.77

 Sergeant

1.49

 Colonel

1.49

 Officer

1.47

Lie

1.47

Activations Density 0.018%

military ranks and titles

instances of the word "Lieutenant" and its abbreviation "Lt."

No Comments

No Known Activations