INDEX

Explanations

references to decision-making and action-oriented phrases

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

asked about

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

 Exact

-0.40

 ***!

-0.37

 архивлан

-0.37

 Administração

-0.37

 znak

-0.35

 stanovnika

-0.35

 objav

-0.35

 camicia

-0.35

 Secrétaire

-0.34

ImageField

-0.34

POSITIVE LOGITS

oneofs

0.61

styleType

0.54

 виправивши

0.54

 '\\;'

0.53

RRect

0.52

 unknownFields

0.51

CloseOperation

0.46

XmlEnum

0.45

minecraftforge

0.45

artist

0.45

Activations Density 0.032%

references to decision-making and action-oriented phrases

asked about

No Comments

No Known Activations