INDEX

Explanations

terms related to methodologies and components in scientific research

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

words consisting of parts

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 AssemblyProduct

-0.43

thenReturn

-0.42

MockBean

-0.41

 propOrder

-0.41

endfor

-0.41

tamo

-0.41

decrement

-0.40

WritableDatabase

-0.40

 Pion

-0.39

 Afri

-0.39

POSITIVE LOGITS

 consisting

0.66

 composed

0.58

 completas

0.57

 completos

0.57

 Zusammensetzung

0.56

 comprised

0.56

 złoż

0.56

 containing

0.55

 completo

0.54

 구성

0.52

Activations Density 0.183%

terms related to methodologies and components in scientific research

words consisting of parts

No Comments

No Known Activations