INDEX

Explanations

references to combat and abilities related to fighting or military actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

 kasarigan

-0.53

sidemargin

-0.41

stdbool

-0.40

 EconPapers

-0.39

UserScript

-0.37

NameInMap

-0.36

普遍

-0.35

Tracce

-0.34

 samfun

-0.34

Initializable

-0.33

POSITIVE LOGITS

 fighting

0.74

 fight

0.71

 attacking

0.67

 fights

0.66

fight

0.65

 fought

0.62

 battle

0.61

 fighters

0.60

 combat

0.59

 battling

0.59

Activations Density 0.335%

references to combat and abilities related to fighting or military actions

No Comments

No Known Activations