INDEX

Explanations

specific character sequences, including punctuation and certain keywords related to identification and naming

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

United States comments

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

 brukar

-0.33

 води

-0.32

translates

-0.29

viez

-0.29

 mariposas

-0.28

 familles

-0.28

 Produktions

-0.28

 alguna

-0.27

 Económica

-0.27

 publiques

-0.26

POSITIVE LOGITS

 CreateTagHelper

0.77

SourceChecksum

0.76

تقاوى

0.76

 nakalista

0.75

 Chwiliwch

0.74

 arşivlendi

0.73

 مرئيه

0.71

setVerticalGroup

0.71

 Савезне

0.69

ArrowToggle

0.69

Activations Density 0.037%

specific character sequences, including punctuation and certain keywords related to identification and naming

United States comments

No Comments

No Known Activations