INDEX

Explanations

conjunctions and prepositions indicating relationships and connections between ideas

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

pairs of words connected by 'and' or 'to'

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

hoeddwyd

-0.42

 amarillas

-0.39

شوف

-0.34

 hábiles

-0.34

 iaitu

-0.31

 gevol

-0.30

 préf

-0.30

الدراسه

-0.30

TestingModule

-0.30

 salvación

-0.30

POSITIVE LOGITS

IntoConstraints

0.76

 AssemblyCulture

0.65

SharedDtor

0.54

 kasarigan

0.53

нгред

0.50

ẖ

0.50

tagHelperRunner

0.48

arsis

0.47

 AssemblyTitle

0.47

 Snapdragon

0.47

Activations Density 0.053%

conjunctions and prepositions indicating relationships and connections between ideas

pairs of words connected by 'and' or 'to'

No Comments

No Known Activations