INDEX

Explanations

symbols or formatting elements often associated with structured data

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

specialized technical terms

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

 للمعارف

-0.37

 camiseta

-0.34

 propia

-0.34

 bolsos

-0.34

 férias

-0.34

 actualidad

-0.34

 bolso

-0.33

 hidratante

-0.33

 kadar

-0.33

 iglesia

-0.33

POSITIVE LOGITS

bootstrapcdn

0.71

expandindo

0.67

ſelves

0.66

parsedMessage

0.64

 Numerade

0.61

WithIOException

0.60

Дереккөздер

0.59

 itſelf

0.59

ſelf

0.58

 Bourgoin

0.58

Activations Density 0.143%

symbols or formatting elements often associated with structured data

specialized technical terms

No Comments

No Known Activations