INDEX

Explanations

method-related keywords and operations in programming contexts

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

verbs related to getting or returning information

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

ThroughAttribute

-0.73

 uLocal

-0.72

 Biôgrafia

-0.71

 الرياضيه

-0.70

-------

-0.69

ſammen

-0.69

 queſta

-0.66

featureID

-0.66

 zwiſchen

-0.66

AsUp

-0.65

POSITIVE LOGITS

the

0.51

 χρή

0.35

 élément

0.35

 selve

0.33

 aplicable

0.33

0.31

 dolay

0.31

 civilización

0.30

등

0.29

Activations Density 0.007%

method-related keywords and operations in programming contexts

verbs related to getting or returning information

No Comments

No Known Activations