INDEX

Explanations

references to news and articles

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

news reporting

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

 للاسماء

-0.42

vmin

-0.36

GRANTED

-0.35

⎠

-0.33

mv

-0.33

utek

-0.33

<eos>

-0.32

Kun

-0.32

nep

-0.31

nps

-0.31

POSITIVE LOGITS

 news

0.85

 noticias

0.76

 News

0.71

 noticia

0.70

 propOrder

0.69

newspaper

0.68

transQ

0.68

news

0.67

News

0.66

reporter

0.66

Activations Density 0.026%

references to news and articles

news reporting

No Comments

No Known Activations