INDEX

Explanations

references to military actions involving Israel and Iran

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

Iran conflict and nuclear weapons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

 bulbs

-0.54

 trimester

-0.53

Enllaços

-0.52

 umbi

-0.51

extAlignment

-0.51

ніципалі

-0.49

 समीक्षाओं

-0.49

 jButton

-0.48

 Whitfield

-0.47

Istorija

-0.47

POSITIVE LOGITS

 McGowan

0.60

 Hezbollah

0.57

versaries

0.56

 Iran

0.54

 Hamas

0.53

Xi

0.52

STEIN

0.51

inhua

0.51

 drones

0.51

伊朗

0.51

Activations Density 0.150%

references to military actions involving Israel and Iran

Iran conflict and nuclear weapons

No Comments

No Known Activations