INDEX

Explanations

phrases indicating improvements, assistance, and positive outcomes

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

enabling positive outcomes

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

 poème

-0.50

hoeddwyd

-0.48

 ModelExpression

-0.47

Parcelize

-0.46

 transfieras

-0.45

IVEREF

-0.45

 Gedicht

-0.45

traversal

-0.45

harmed

-0.45

 disrupting

-0.44

POSITIVE LOGITS

EndContext

0.42

MeasureSpec

0.34

 magasiner

0.34

AndEndTag

0.34

}{*}{

0.33

 concor

0.31

sml

0.30

 상세

0.30

ffions

0.30

cipital

0.30

Activations Density 0.117%

phrases indicating improvements, assistance, and positive outcomes

enabling positive outcomes

No Comments

No Known Activations