INDEX

Explanations

concepts related to information gathering and assessment of its usefulness

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 kasarigan

-0.60

parsedMessage

-0.57

 pinulongan

-0.54

thâu

-0.53

 للمعارف

-0.52

Personendaten

-0.51

 iNdEx

-0.50

akhstan

-0.48

 ТАК

-0.48

 للاسماء

-0.47

POSITIVE LOGITS

出版年

0.42

PreExecute

0.42

IActionResult

0.37

 unnecessarily

0.36

WithFormat

0.35

そりゃ

0.34

conio

0.34

 needlessly

0.33

ungeon

0.32

 Gebä

0.32

Activations Density 0.097%

concepts related to information gathering and assessment of its usefulness

No Comments

No Known Activations