INDEX

Explanations

phrases indicating the presence or actions of people and their relationships to certain events or circumstances

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

Followed by action/modality words ("being", "able", "forced")

np_acts-logits-general · gemini-2.0-flash

states or conditions

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

UpInside

-0.49

ims

-0.48

nologies

-0.48

otka

-0.46

 главе

-0.46

 begleiten

-0.45

DllImport

-0.45

 sự

-0.44

 flagship

-0.44

yoto

-0.44

POSITIVE LOGITS

 forced

0.76

 flock

0.69

 ausgeschlossen

0.67

 encouraged

0.67

 clamor

0.66

 estimés

0.63

 otomatig

0.63

 finding

0.63

forced

0.62

 createSlice

0.62

Activations Density 0.255%

phrases indicating the presence or actions of people and their relationships to certain events or circumstances

Followed by action/modality words ("being", "able", "forced")

states or conditions

No Comments

No Known Activations

phrases indicating the presence or actions of people and their relationships to certain events or circumstances

Followed by action/modality words ("being", "able", "forced")

states or conditions

No Comments

No Known Activations