INDEX

Explanations

programming-related terms and functions

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

directory paths or names

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ToTensor

-0.41

Referencer

-0.33

vigil

-0.32

 György

-0.31

 sánh

-0.29

 ModelExpression

-0.29

vig

-0.28

volley

-0.27

 exer

-0.27

 дописавши

-0.27

POSITIVE LOGITS

 directories

0.76

 directory

0.75

 dirs

0.75

isdir

0.74

 Directory

0.72

DIRECTORY

0.71

dirs

0.70

directories

0.67

 DIRECTORY

0.66

directory

0.66

Activations Density 0.054%

programming-related terms and functions

directory paths or names

No Comments

No Known Activations