INDEX

Explanations

terms related to the concept of correctness and appropriateness

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

"Right" or "more" followed by a noun/pronoun

np_acts-logits-general · gemini-2.0-flash

more right wrong

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.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

 arşivlendi

-0.42

 GoogleFonts

-0.38

NameInMap

-0.38

DEPORTES

-0.35

してみ

-0.32

Parcelize

-0.32

 CrossRef

-0.31

GEBURTSDATUM

-0.31

jangkau

-0.30

AISSEE

-0.30

POSITIVE LOGITS

HideFlags

0.68

ConstraintMaker

0.68

CloseOperation

0.58

setVerticalGroup

0.57

 thing

0.57

findpost

0.57

 snippetHide

0.56

interopRequire

0.55

Handlung

0.54

fjspx

0.53

Activations Density 0.807%

terms related to the concept of correctness and appropriateness

"Right" or "more" followed by a noun/pronoun

more right wrong

No Comments

No Known Activations