INDEX

Explanations

phrases associated with violence and conflict

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

attacking

np_acts-logits-general · gemini-2.0-flash

various spells

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

fjspx

-0.66

مصادر

-0.65

 يتيمه

-0.59

 kaarangay

-0.59

 VIAF

-0.56

iov

-0.51

 محفوظة

-0.50

 kurios

-0.50

ParallelGroup

-0.49

 tjen

-0.49

POSITIVE LOGITS

 attacking

0.75

TagMode

0.75

 attack

0.69

attack

0.65

 attacked

0.65

 attacks

0.63

Attacking

0.61

 MainAxisSize

0.60

 shot

0.60

 besie

0.59

Activations Density 0.055%

phrases associated with violence and conflict

attacking

various spells

No Comments

No Known Activations