INDEX

Explanations

content related to conclusions or summaries in texts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

in particular

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

Negative Logits

 nakalista

-0.82

 informée

-0.69

TagMode

-0.65

 CanadaChoose

-0.65

tvguidetime

-0.63

Tembelea

-0.63

 surla

-0.63

تقاوى

-0.61

RectangleBorder

-0.60

 ſind

-0.59

POSITIVE LOGITS

 teş

0.30

astify

0.30

 biais

0.29

 keper

0.29

 Erscheinung

0.28

 politie

0.28

น์

0.28

 zdrav

0.27

 veroor

0.27

dagangan

0.27

Activations Density 0.055%

content related to conclusions or summaries in texts

in particular

No Known Activations