INDEX

Explanations

terms related to claims and proposals

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

proposed or alleged

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

 člověk

-0.50

bø

-0.44

 jorden

-0.42

 cuchara

-0.41

 anybody

-0.41

 auroit

-0.40

entada

-0.40

 alguém

-0.39

 dalamnya

-0.39

 conmigo

-0.39

POSITIVE LOGITS

تقاوى

0.71

OGND

0.69

WriteTagHelper

0.69

__*/

0.67

 Desired

0.65

cipated

0.61

 Planned

0.60

 ویکی‌پدیای

0.60

Życiorys

0.59

Desired

0.59

Activations Density 0.056%

terms related to claims and proposals

proposed or alleged

No Comments

No Known Activations

terms related to claims and proposals

proposed or alleged

No Comments

No Known Activations