INDEX

Explanations

references to specific media and health-related topics

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

pairs of entities

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

новниш

-0.51

mobileqq

-0.50

 ErrIntOverflow

-0.47

󠁴

-0.46

 dAtA

-0.46

setVerticalGroup

-0.43

contentLoaded

-0.43

 TextAppearance

-0.42

 Numerade

-0.41

 FIFTH

-0.40

POSITIVE LOGITS

 both

0.56

IntoConstraints

0.54

 alike

0.48

 beider

0.48

 begge

0.46

 entrambi

0.46

 beiden

0.45

both

0.45

 ambos

0.45

 beide

0.43

Activations Density 0.098%

references to specific media and health-related topics

pairs of entities

No Comments

No Known Activations