INDEX

Explanations

terms related to data structure manipulation and transformation

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

flattening data structures or hierarchies

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

+#+#

-0.49

 ब्रेकडाउन

-0.45

EReference

-0.44

 <>",

-0.42

存于互联网档案馆

-0.42

[]:

-0.41

+:+

-0.41

fjspx

-0.41

Ɓ

-0.40

rungsseite

-0.40

POSITIVE LOGITS

flatten

0.93

Flatten

0.90

 Flatten

0.80

 flatten

0.79

 flattening

0.71

 flattened

0.65

 GoogleFonts

0.54

 sederhana

0.53

ChildScrollView

0.53

internalType

0.52

Activations Density 0.017%

terms related to data structure manipulation and transformation

flattening data structures or hierarchies

No Comments

No Known Activations