INDEX

Explanations

first-person pronouns and expressions of self-reference

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

:✨

-0.57

Personally

-0.50

 personally

-0.49

 personalmente

-0.48

 Personally

-0.47

 myself

-0.42

ChildScrollView

-0.42

 Obec

-0.41

 ourselves

-0.40

 View

-0.39

POSITIVE LOGITS

 better

0.54

 guess

0.54

better

0.52

 SAID

0.50

enderror

0.50

 AssemblyCulture

0.46

 suppose

0.46

 meant

0.45

 חיצוניים

0.44

 warned

0.44

Activations Density 0.212%

first-person pronouns and expressions of self-reference

No Comments

No Known Activations