INDEX

Explanations

references to notable individuals and their actions or statements

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

when

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

 ujednoznacz

-0.63

tagHelperRunner

-0.54

 defaultstate

-0.52

rungsseite

-0.48

 snippetHide

-0.46

ValueStyle

-0.45

enumii

-0.45

principalTable

-0.45

beyond

-0.45

 sebou

-0.43

POSITIVE LOGITS

 when

0.67

Когда

0.64

Quand

0.64

Quando

0.63

Cuando

0.63

When

0.62

Cuándo

0.62

 WHEN

0.60

 quando

0.60

 When

0.60

Activations Density 0.054%

references to notable individuals and their actions or statements

when

No Comments

No Known Activations