INDEX

Explanations

dates and numerical references related to months and years

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

names of months

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

uxx

-0.52

 nakalista

-0.47

GraphicsUnit

-0.46

タッチ

-0.46

 saites

-0.45

+#+

-0.45

utilisons

-0.44

 Touch

-0.44

AutoField

-0.43

béco

-0.42

POSITIVE LOGITS

GEBURTSDATUM

0.52

 month

0.40

 ujednoznacz

0.39

TypedDataSet

0.38

 busiest

0.38

 Month

0.37

OOTDTY

0.36

Aza

0.34

BeginInit

0.34

 minggu

0.34

Activations Density 0.070%

dates and numerical references related to months and years

names of months

No Comments

No Known Activations