INDEX

Explanations

references to existing research or methods in the context of scientific discussions

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

technical processing and devices

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

Plo

-0.35

NUMX

-0.34

Prompt

-0.33

 prompt

-0.32

 répar

-0.30

屯

-0.30

trä

-0.29

立

-0.28

 Prompt

-0.28

 habet

-0.27

POSITIVE LOGITS

RenderAtEndOf

0.72

FormTagHelper

0.54

KommentareTeilen

0.54

 ब्रेकडाउन

0.53

 informée

0.53

 INTERESAR

0.50

tagHelperRunner

0.50

+#+#

0.50

orteur

0.48

 betweenstory

0.48

Activations Density 0.086%

references to existing research or methods in the context of scientific discussions

technical processing and devices

No Comments

No Known Activations