INDEX

Explanations

important keywords and phrases related to specificity and action in various contexts

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

boys and girls preferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RegistryLite

-0.70

IntoConstraints

-0.65

DockStyle

-0.63

ToBounds

-0.61

Хьажоргаш

-0.59

archiviato

-0.57

astify

-0.57

 esternos

-0.56

PYX

-0.56

addPreferredGap

-0.54

POSITIVE LOGITS

CrossRef

0.42

Netflix

0.36

])

0.34

böz

0.33

severity

0.33

Koordin

0.32

Sib

0.32

 Netflix

0.31

ilu

0.31

完全

0.31

Activations Density 0.001%

important keywords and phrases related to specificity and action in various contexts

boys and girls preferences

No Comments

No Known Activations