INDEX

Explanations

phrases indicating recommendations or high levels of appreciation

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

high praise, recommendations, and positive remarks

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

tagHelperRunner

-0.82

 विश्वसनीयता

-0.79

DockStyle

-0.79

 المعيارى

-0.77

Vidite

-0.76

 transfieras

-0.76

ValueStyle

-0.75

تقاوى

-0.75

 חיצוניים

-0.73

AndEndTag

-0.73

POSITIVE LOGITS

 excellent

0.79

 Excellent

0.73

 wonderful

0.69

 kudos

0.69

 great

0.68

Excellent

0.68

 👍

0.67

 definitely

0.66

 staff

0.66

 excelente

0.65

Activations Density 0.010%

phrases indicating recommendations or high levels of appreciation

high praise, recommendations, and positive remarks

No Comments

No Known Activations