INDEX

Explanations

references to individuals and groups as subjects in various contexts

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

like "who" or question words

np_acts-logits-general · gemini-2.0-flash

who are or may find

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.30.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

awtextra

-0.68

itself

-0.61

出版年

-0.60

RTLR

-0.59

 initComponents

-0.54

 itself

-0.52

 surla

-0.51

SuspendLayout

-0.50

 Itself

-0.50

ItemBackground

-0.48

POSITIVE LOGITS

who

0.65

 którzy

0.49

 whom

0.44

Izvori

0.42

 الذين

0.42

 innocently

0.40

 speculating

0.40

kto

0.40

 ktorí

0.40

argento

0.39

Activations Density 0.433%

references to individuals and groups as subjects in various contexts

like "who" or question words

who are or may find

No Comments

No Known Activations