INDEX

Explanations

terms and phrases related to programming concepts, particularly those involving function definitions, parameters, and data handling

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

followed by prepositions

np_acts-logits-general · gemini-2.0-flash

verbs ending in -ed or -ing

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 folhas

-0.40

 avoient

-0.38

citenamefont

-0.38

guém

-0.38

AddHtmlAttribute

-0.37

 wią

-0.35

 SqlParameter

-0.35

 flash

-0.34

 InputDecoration

-0.34

 Superhost

-0.33

POSITIVE LOGITS

 cherchés

0.57

 censiti

0.48



0.48

RTDA

0.47

rungsseite

0.44

parsedMessage

0.44

Lähteet

0.43

ंदीखरीदारी

0.43

 propOrder

0.43

 <<<<<<<<<<<<<<

0.43

Activations Density 0.597%

terms and phrases related to programming concepts, particularly those involving function definitions, parameters, and data handling

followed by prepositions

verbs ending in -ed or -ing

No Comments

No Known Activations