INDEX

Explanations

coding-related syntax and structure

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

increment

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 betweenstory

-0.39

cisme

-0.30

uxxxx

-0.30

reathable

-0.30

 thiệu

-0.29

httphttps

-0.29

 Besøkt

-0.29

 långt

-0.29

resave

-0.28

คุ

-0.28

POSITIVE LOGITS

 increment

1.68

 increase

1.68

 incre

1.57

 Increase

1.53

increase

1.51

 Increment

1.47

Increase

1.47

 increases

1.45

 INCRE

1.44

 INCREASE

1.42

Activations Density 0.057%

coding-related syntax and structure

increment

No Known Activations