INDEX

Explanations

negative expressions or negations

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

negation followed by descriptive word

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

Grüne

-0.46

DDD

-0.45

 Trä

-0.45

using

-0.43

apps

-0.43

ideas

-0.43

fél

-0.42

☆☆☆☆☆

-0.42

HJ

-0.42

idea

-0.42

POSITIVE LOGITS

not

1.02

not

0.85

Not

0.84

NOT

0.83

NotBe

0.77

Tidak

0.77

IsNot

0.75

Not

0.74

 necessarily

0.73

isNot

0.72

Activations Density 0.142%

negative expressions or negations

negation followed by descriptive word

No Comments

No Known Activations

negative expressions or negations

negation followed by descriptive word

No Comments

No Known Activations