INDEX

Explanations

references to events, actions, or occurrences related to competition or ranking

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

from, since, before, survey

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

 autorytatywna

-0.46

userProfile

-0.41

 axes

-0.40

put

-0.38

 onStop

-0.35

|};

-0.35

 Vidal

-0.33

feld

-0.33

 acid

-0.33

gang

-0.33

POSITIVE LOGITS

DockStyle

0.82

 виправивши

0.63

AnchorStyles

0.63

IntoConstraints

0.59

centralwidget

0.57

MLLoader

0.56

belea

0.55

 linkovi

0.54

expandindo

0.54

 <<<<<<<<<<<<<<

0.54

Activations Density 0.482%

references to events, actions, or occurrences related to competition or ranking

from, since, before, survey

No Comments

No Known Activations