INDEX

Explanations

references to ships and related terminology

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

ships and maritime contexts

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Kendal

-0.58

 Dani

-0.56

QT

-0.53

liven

-0.52

queryInterface

-0.51

orama

-0.51

 futile

-0.51

隈

-0.50

 Lila

-0.49

ondere

-0.49

POSITIVE LOGITS

Ship

1.09

 ship

1.05

 Ship

1.05

Ships

1.01

 Ships

1.01

 ships

0.96

 SHIP

0.92

 navire

0.87

ship

0.86

ships

0.77

Activations Density 0.007%

references to ships and related terminology

ships and maritime contexts

No Comments

No Known Activations