INDEX

Explanations

sequences of code-related syntax elements and constructs

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

method calls or variable access

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

AddTagHelper

-0.56

principalColumn

-0.47

 للاسماء

-0.42

 nahilalakip

-0.40

 الرياضيه

-0.40

Gön

-0.39

AutoField

-0.38

 <<<<<<<<<<<<<<

-0.38

 invokingState

-0.38

 <=",

-0.38

POSITIVE LOGITS

rungsseite

0.50

Notae

0.45

loroethane

0.40

erebbe

0.39

eneuve

0.39

GMA

0.38

Garten

0.38

parsedMessage

0.38

resolvers

0.38

degna

0.38

Activations Density 0.048%

sequences of code-related syntax elements and constructs

method calls or variable access

No Comments

No Known Activations