INDEX

Explanations

terms related to software or computing processes

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

interceptor or capacitor

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

TypedDataSet

-0.40

躇

-0.38

<![

-0.37

<h2>

-0.37

 Putih

-0.36

 berjudul

-0.36

inginan

-0.35

POSITIVE LOGITS

Interceptor

1.99

interceptors

1.52

 Interceptor

1.38

 interceptor

1.38

interceptor

1.38

ceptors

0.83

 interception

0.79

拦截

0.77

intercept

0.74

 intercept

0.73

Activations Density 0.001%

terms related to software or computing processes

interceptor or capacitor

No Comments

No Known Activations