INDEX

Explanations

forms of the verb "to be."

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

states and descriptions

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

Климат

-0.54

 kasarigan

-0.51

featureID

-0.50

 ब्रेकडाउन

-0.46

Rcpp

-0.45

 itſelf

-0.44

éez

-0.42

QtGui

-0.42

RegistryLite

-0.42

 TextAppearance

-0.42

POSITIVE LOGITS

ontale

0.43

 Christiansen

0.42

 Persson

0.42

 equipments

0.41

 coté

0.40

 satunya

0.40

omencl

0.39

 Herrmann

0.38

etc

0.38

 Schulze

0.38

Activations Density 0.046%

forms of the verb "to be."

states and descriptions

No Comments

No Known Activations