INDEX

Explanations

phrases or terms related to legal or formal documentation

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

technical programming terms

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

 CURIAM

-0.46

KindOfClass

-0.38

 يتيمه

-0.38

contentLoaded

-0.37

rungsseite

-0.37

AccessorTable

-0.36

BeforeAll

-0.35

المناصب

-0.34

 miliar

-0.34

POSITIVE LOGITS

Datuak

0.50

0.47

 referrerpolicy

0.47

extAlignment

0.46

aarrggbb

0.41

 lenker

0.41

 AssemblyProduct

0.41

MarshalTo

0.40

 рабочего

0.40

 "..\..\

0.39

Activations Density 0.148%

phrases or terms related to legal or formal documentation

technical programming terms

No Comments

No Known Activations