INDEX

Explanations

numerical references to options or methods in a procedural context

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

options, alternatives, second methods

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

 demografica

-0.50

 Reap

-0.50

HasAnnotation

-0.49

LEncoder

-0.48

ooplankton

-0.47

 Longueur

-0.47

 Offisielt

-0.47

éfonos

-0.47

identiality

-0.45

 ubezpiec

-0.45

POSITIVE LOGITS

 alternatively

0.42

 Alternatively

0.41

Alternatively

0.41

 simply

0.38

 option

0.38

option

0.38

 lick

0.38

Suppose

0.36

 assume

0.36

 Möglichkeit

0.36

Activations Density 0.082%

numerical references to options or methods in a procedural context

options, alternatives, second methods

No Comments

No Known Activations