INDEX

Explanations

indicators of emotions or states of being in various contexts

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

tens of thousands

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

 entendre

-0.38

audiovisuel

-0.35

 Trä

-0.34

 graça

-0.34

 Truly

-0.33

ьаж

-0.33

ouml

-0.33

 gustaría

-0.32

Dog

-0.32

iál

-0.32

POSITIVE LOGITS

extAlignment

0.54

AnimationsModule

0.54

下载附件

0.49

évaluateur

0.46

MethodManager

0.45

原始内容存档于

0.43

 ProtoMessage

0.43

rungsseite

0.43

mergeFrom

0.43

 StatefulWidget

0.42

Activations Density 0.051%

indicators of emotions or states of being in various contexts

tens of thousands

No Comments

No Known Activations