INDEX

Explanations

phrases indicating necessity or obligation

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

must do

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

 vlieg

-0.38

RegressionTest

-0.37

Vorte

-0.35

 зви

-0.33

 Völ

-0.32

 idéia

-0.32

ilang

-0.32

typeName

-0.31

 proyec

-0.31

 clasificación

-0.30

POSITIVE LOGITS

Must

0.90

 Must

0.90

 must

0.85

 Нужно

0.85

 Gotta

0.82

 gotta

0.81

 Надо

0.79

MUST

0.78

 MUST

0.78

是要

0.78

Activations Density 0.023%

phrases indicating necessity or obligation

must do

No Comments

No Known Activations