INDEX

Explanations

phrasings related to preference or prioritization

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

references to giving preference, priority, or favor to something or someone.

oai_token-act-pair · gpt-4.1-2025-04-14 Triggered by @ihobbang250

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 CGSize

-0.54

 initComponents

-0.54

gbaar

-0.52

 alrededores

-0.50

sums

-0.50

 nuages

-0.49

 coster

-0.48

ETHING

-0.47

์ตูน

-0.47

 inex

-0.47

POSITIVE LOGITS

 prefer

0.85

prefer

0.77

 preference

0.77

 preferred

0.75

 Prefer

0.71

 preferential

0.69

preferred

0.69

 Preference

0.68

 preferentially

0.67

 prioritize

0.65

Activations Density 0.050%

phrasings related to preference or prioritization

references to giving preference, priority, or favor to something or someone.

No Comments

No Known Activations