INDEX

Explanations

negative values in mathematical expressions or discussions

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

methadone maintenance

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

 indígen

-0.81

 pinulongan

-0.80

 ſind

-0.76

 niedersachsen

-0.76

tvguidetime

-0.75

 ddelwed

-0.75

 TextInputType

-0.75

 Infór

-0.74

يكب

-0.72

 يتيمه

-0.72

POSITIVE LOGITS

0.89

item

0.77

0.71

•

0.65

0.62

•

0.60

cdot

0.60

>-</

0.58

0.56

>−</

0.56

Activations Density 0.000%

negative values in mathematical expressions or discussions

methadone maintenance

No Comments

No Known Activations