INDEX

Explanations

instances of the letter 'e' and certain function calls or log statements in code

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

Android logs and code

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

ロウィン

-0.61

leria

-0.61

__((

-0.60

 للاسماء

-0.60

picasso

-0.58

isContained

-0.57

$_-

-0.57

HtmlAttribute

-0.56

 ſind

-0.56

INSEE

-0.56

POSITIVE LOGITS

 telefónica

0.38

 rayas

0.38

 transférez

0.38

 venganza

0.36

 Revolución

0.35

 civilización

0.35

texttt

0.35

 bahía

0.35

 visuales

0.35

 exportaciones

0.34

Activations Density 0.002%

instances of the letter 'e' and certain function calls or log statements in code

Android logs and code

No Comments

No Known Activations