INDEX

Explanations

references to subwoofers or related audio equipment

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

sub- prefix

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

<bos>

-0.55

glers

-0.43

 Opportun

-0.42

 ações

-0.41

 mắn

-0.41

neux

-0.41

assure

-0.41

ADEM

-0.41

ksis

-0.40

abilities

-0.40

POSITIVE LOGITS

SUB

1.20

Sub

1.18

sub

1.17

SUB

1.13

Sub

1.12

 getSub

1.04

 subs

1.02

sub

1.01

 Subs

0.97

Subs

0.93

Activations Density 0.021%

references to subwoofers or related audio equipment

sub- prefix

No Comments

No Known Activations