INDEX

Explanations

phrases indicating code generation and its origins

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

code generated by

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

lifts

-0.39

стви

-0.38

GOTREF

-0.38

RetentionPolicy

-0.36

 Creator

-0.36



-0.36

nodoc

-0.35

__()

-0.35

WebServlet

-0.35

 especially

-0.34

POSITIVE LOGITS

IUrlHelper

0.63

RTDA

0.62

Tikang

0.56

uxxxx

0.56

CppCodeGen

0.53

elemField

0.52

Autoritní

0.52

 CanadaChoose

0.52

 invokingState

0.50

AutoScaleMode

0.50

Activations Density 0.006%

phrases indicating code generation and its origins

code generated by

No Comments

No Known Activations