INDEX

Explanations

terms related to randomness and selection processes in data manipulation

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

randomizing and shuffling

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

RTEX

-0.56

 InputDecoration

-0.54

setVerticalGroup

-0.54

exitRule

-0.53

PreferredItem

-0.52

MemoryWarning

-0.51

 desmotivaciones

-0.51

olerance

-0.51

ScopeManager

-0.50

 TextAppearance

-0.50

POSITIVE LOGITS

shuffle

0.49

 randomly

0.45

 random

0.43

invokeLater

0.40

 shuffled

0.40

Random

0.37

 Random

0.37

shuff

0.36

srand

0.36

 rand

0.35

Activations Density 0.006%

terms related to randomness and selection processes in data manipulation

randomizing and shuffling

No Comments

No Known Activations