INDEX

Explanations

punctuation marks and formatting elements indicative of dialogue or script-like content

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

code and technical terms

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

Negative Logits

expandindo

-0.56

 florales

-0.46

 nød

-0.41

 ótima

-0.41

 tradizione

-0.40

ChromeDriver

-0.40

 chré

-0.39

 tradición

-0.38

 sienta

-0.38

 agrícolas

-0.38

POSITIVE LOGITS

httphttps

0.55

DeleteBehavior

0.52

CppMethod

0.52

UNIDENTIFIED

0.51

UnusedPrivate

0.49

 defaultstate

0.47

PerformLayout

0.44

bitField

0.42

magnitude

0.41

SOUNDBITE

0.41

Activations Density 0.067%

punctuation marks and formatting elements indicative of dialogue or script-like content

code and technical terms

No Known Activations