INDEX

Explanations

specific geographical locations or proper nouns, particularly names of places

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

proper nouns

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

 transfieras

-0.78

OGND

-0.68

 propOrder

-0.66

 дописавши

-0.65

Diweddarwch

-0.60

AnchorTagHelper

-0.59

Autoritní

-0.58

 Italijanski

-0.57

новниш

-0.57

 unknownFields

-0.56

POSITIVE LOGITS

 parachoque

0.47

RegressionTest

0.42

 esferas

0.38

 suceso

0.36

 duquel

0.36

ースター

0.33

 observaciones

0.32

 inilah

0.32

 ocasión

0.31

 Schule

0.31

Activations Density 0.035%

specific geographical locations or proper nouns, particularly names of places

proper nouns

No Comments

No Known Activations