INDEX

Explanations

military-related timeframes and designations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

MLLoader

-0.76

ьаж

-0.63

 snippetHide

-0.61

ьажоргаш

-0.61

IntOverflow

-0.60

 autorytatywna

-0.59

 SwitchCompat

-0.59

 gynhyrchwyd

-0.58

sizeCache

-0.58

 Exacts

-0.57

POSITIVE LOGITS

bezpiecz

0.41

 dangereux

0.36

 dangerous

0.35

 nocturno

0.32

 preventiva

0.32

Dangerous

0.32

 parti

0.31

 receb

0.30

 możliwe

0.30

 pelig

0.30

Activations Density 0.036%

military-related timeframes and designations

No Comments

No Known Activations