INDEX

Explanations

names of royalty or historical figures

oai_token-act-pair · gpt-3.5-turbo

titles and ranks associated with nobility and historical figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

paces

-0.96

roups

-0.82

 layoffs

-0.79

groups

-0.79

Apps

-0.79

 salads

-0.77

tones

-0.77

poons

-0.76

rooms

-0.76

 clusters

-0.75

POSITIVE LOGITS

 ruler

1.59

 emperor

1.56

 Emperor

1.54

 commander

1.47

 dictator

1.45

 Empress

1.38

 tyrant

1.35

 Ruler

1.34

 Commander

1.30

 governor

1.30

Activations Density 0.341%

names of royalty or historical figures

titles and ranks associated with nobility and historical figures

No Comments

No Known Activations