INDEX

Explanations

phrases indicating predictions or projections

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

by [time]

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

 gustan

-0.49

 miniaturka

-0.46

 cucharadita

-0.46

zoeken

-0.42

 sillas

-0.40

ศาสตร์

-0.40

eraard

-0.40

 maschile

-0.40

 älskar

-0.40

lijck

-0.40

POSITIVE LOGITS

明年

0.60

 next

0.59

 tomorrow

0.58

 оригіналу

0.57

来年

0.55

DockStyle

0.55

 sometime

0.54

 early

0.52

 future

0.50

 until

0.50

Activations Density 0.015%

phrases indicating predictions or projections

by [time]

No Comments

No Known Activations