INDEX

Explanations

references or links to additional resources or information

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

link to

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

stylers

-0.50

 Infórmanos

-0.50

Rüyada

-0.46

TacToe

-0.45

ControllerAdvice

-0.45

Aholisi

-0.44

SystemColors

-0.44

__":

-0.42

 séries

-0.41

styleType

-0.41

POSITIVE LOGITS

 link

0.89

 links

0.88

link

0.72

links

0.71

 Links

0.68

 LINK

0.67

Links

0.66

 Link

0.65

LINK

0.65

Link

0.64

Activations Density 0.018%

references or links to additional resources or information

link to

No Comments

No Known Activations