INDEX

Explanations

terms related to buffer overflow vulnerabilities in software

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

user input signals

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

ThroughAttribute

-0.41

 tươi

-0.39

 universelle

-0.38

jooq

-0.38

ztes

-0.37

 citoy

-0.36

 impressions

-0.36

 touristique

-0.35

klis

-0.35

 comis

-0.34

POSITIVE LOGITS

 Vikipedi

1.02

usercontent

0.97

 consectetur

0.86

="+

0.83

 ***!

0.80

 betweenstory

0.75

BeginContext

0.67

ArrowToggle

0.65

 estekak

0.59

enterOuterAlt

0.59

Activations Density 0.001%

terms related to buffer overflow vulnerabilities in software

user input signals

No Comments

No Known Activations