INDEX

Explanations

date and time references

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

months and day/year numbers

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

 comienzo

-0.34

 apropiado

-0.34

 apertura

-0.33

ormick

-0.33

 perfeito

-0.33

所で

-0.33

̀u

-0.31

ynn

-0.30

diques

-0.30

 zaczyna

-0.30

POSITIVE LOGITS

 until

0.65

GEBURTSDATUM

0.62

WaitGroup

0.61

until

0.58

 Until

0.57

Until

0.57

 <<<<<<<<<<<<<<

0.57

EDEFAULT

0.56

tagHelperRunner

0.56

ChildScrollView

0.55

Activations Density 0.020%

date and time references

months and day/year numbers

No Comments

No Known Activations