INDEX

Explanations

terms related to surveillance, targeting, and various forms of attack or conflict

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

іга

-0.41

使其

-0.36

lī

-0.35

 efectivos

-0.34

 húmedo

-0.34

verwijspagina

-0.34

 installé

-0.32

 lumineuse

-0.32

 cálido

-0.32

 पता

-0.32

POSITIVE LOGITS

 gynhyrchwyd

0.59

 targets

0.57

 cible

0.54

 Targets

0.52

AndroidJUnit

0.52

targets

0.51

 enemies

0.50

 enemy

0.49

onOptions

0.49

 Jefus

0.48

Activations Density 1.078%

terms related to surveillance, targeting, and various forms of attack or conflict

No Comments

No Known Activations