INDEX

Explanations

specific numerical and relational data, often in the context of comparisons or statistics

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

positions above, spat, Thornton had

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

seamnă

-0.41

AnchorStyles

-0.41

MENAFN

-0.38

 yüzden

-0.36

getSystemService

-0.34

 pungkas

-0.33

 preuves

-0.33

 Consumo

-0.32

vignon

-0.32

Билгалдахарш

-0.32

POSITIVE LOGITS

Географиясе

0.60

EndGlobal

0.52

 للاسماء

0.51

ftagPool

0.49

 Ordinary

0.47

 inexperienced

0.46

styleType

0.44

 Tenggara

0.43

 getSource

0.43

 humble

0.42

Activations Density 0.108%

specific numerical and relational data, often in the context of comparisons or statistics

positions above, spat, Thornton had

No Comments

No Known Activations