INDEX

Explanations

commands or requests related to actions in a programming or technical context

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

action verbs on data

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

imperative, task-directing commands at the start of a prompt or instruction line.

oai_token-act-pair · gpt-5 Triggered by @vetterc0

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

Попис

-0.73

:✨

-0.72

 bezeichneter

-0.65

RTEE

-0.65

audiovisuel

-0.65

 Мексичка

-0.64

Rüyada

-0.64

SharedDtor

-0.63

WebElementEntity

-0.62

⅘

-0.61

POSITIVE LOGITS

ValueStyle

0.37

expandindo

0.35

mathrm

0.34

 myself

0.33

 Quelles

0.32

ExtendWith

0.30

を渡

0.29

minen

0.29

 levier

0.29

 henne

0.28

Activations Density 0.040%

commands or requests related to actions in a programming or technical context

action verbs on data

imperative, task-directing commands at the start of a prompt or instruction line.

No Comments

No Known Activations

commands or requests related to actions in a programming or technical context

action verbs on data

imperative, task-directing commands at the start of a prompt or instruction line.

No Comments

No Known Activations