INDEX

Explanations

names of characters and their relationships, particularly focusing on family members

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

同学

-0.35

 otomatig

-0.32

 leid

-0.31

experimental

-0.30

 angst

-0.30

生

-0.30

 ardour

-0.30

同學

-0.29

 Administrativna

-0.29

gmx

-0.29

POSITIVE LOGITS

👵

0.96

 Elderly

0.92

 elderly

0.88

👴

0.84

 retired

0.83

 ancianos

0.82

 grandfather

0.81

 grandmother

0.80

 Grandmother

0.80

 Grandfather

0.79

Activations Density 0.385%

names of characters and their relationships, particularly focusing on family members

No Comments

No Known Activations