INDEX

Explanations

patterns or structures in visual or tabular data representations

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

table structure and formatting

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

Notae

-0.51

 custodian

-0.49

URLException

-0.47

aulted

-0.45

vább

-0.45

 snippetHide

-0.43

Vanjske

-0.42

 custodi

-0.42

Captor

-0.41

EndGlobal

-0.41

POSITIVE LOGITS

ContentAsync

0.50

AsUp

0.48

Gön

0.41

 kesk

0.41

webElementXpaths

0.39

BASEPATH

0.39

cinhas

0.39

 Trennung

0.38

 preocu

0.37

ParallelGroup

0.37

Activations Density 0.002%

patterns or structures in visual or tabular data representations

table structure and formatting

No Comments

No Known Activations