INDEX

Explanations

references to forms, objects, or structures within programming or database contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

numbers and quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

évaluateur

-0.52

 pinulongan

-0.52



-0.52

 kasarigan

-0.49

 yourself

-0.48

Autoritní

-0.47

Personendaten

-0.47

了吧

-0.42

탤

-0.41

 Puoi

-0.40

POSITIVE LOGITS

 several

0.67

 hundreds

0.63

several

0.60

 some

0.60

 plusieurs

0.60

 occasionally

0.60

 beberapa

0.60

 GenerationType

0.60

 dozens

0.59

 sometimes

0.59

Activations Density 0.406%

references to forms, objects, or structures within programming or database contexts

numbers and quantities

No Known Activations