INDEX

Explanations

the word "no" in various contexts, particularly emphasizing negation or absence

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

no + descriptor

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

 évêque

-0.40

 Putih

-0.37

 bağlı

-0.35

 temiz

-0.35

 jeweils

-0.34

with

-0.34

 rast

-0.34

 עם

-0.33

Id

-0.33

The

-0.32

POSITIVE LOGITS

ſcher

0.73

ſehen

0.70

niſſe

0.69

iſchen

0.69

 AssemblyTitle

0.67

<unused16>

0.66

enablog

0.66

[@BOS@]

0.66

<unused17>

0.65

<unused41>

0.65

Activations Density 0.012%

the word "no" in various contexts, particularly emphasizing negation or absence

no + descriptor

No Comments

No Known Activations