INDEX

Explanations

dates and temporal references related to events

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

years and counts

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

@"/

-0.63

RectangleBorder

-0.40

IntoConstraints

-0.40

ButterKnife

-0.38

RenderAtEndOf

-0.37

.*")]

-0.36

توض

-0.35

 createState

-0.35

LEncoder

-0.34

 jusque

-0.34

POSITIVE LOGITS

varande

0.49

fein

0.46

يكب

0.46

EndGlobalSection

0.45

Unset

0.44

konom

0.44

ifflin

0.44

Géographie

0.43

ocs

0.43

numbered

0.43

Activations Density 0.024%

dates and temporal references related to events

years and counts

No Comments

No Known Activations