INDEX

Explanations

regular expressions and patterns in text, specifically those related to matching and capturing groups

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

match regex

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

__':

-0.56

__':

-0.52

 MessageBoxIcon

-0.50

]")]

-0.50

__":

-0.49

__":

-0.48

 heapq

-0.45

addGroup

-0.44

 pleaſure

-0.44

 NgModule

-0.43

POSITIVE LOGITS

 Савезне

0.77

 else

0.53

EndInit

0.50

(.*

0.50

 تضيفلها

0.50

WriteBarrier

0.50

 contenido

0.47

 OFDb

0.47

 Abril

0.46

 fleste

0.45

Activations Density 0.020%

regular expressions and patterns in text, specifically those related to matching and capturing groups

match regex

No Comments

No Known Activations