INDEX

Explanations

specific codes or identifiers likely related to software or programming contexts

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

hexadecimal strings and bytes

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

 Signalez

-0.68

ValueStyle

-0.59

RTLR

-0.56

MessageOf

-0.55

}{*}{

-0.54

нгред

-0.53

 désolés

-0.53

RTLD

-0.52

}")]

-0.51

 témoig

-0.50

POSITIVE LOGITS

Халык

0.41

 Aragon

0.38

 Normdatei

0.38

 SwitchCompat

0.36

 ویکی‌پدی

0.36

 ujarnya

0.36

bcryptjs

0.35

ManagedBean

0.33

XtraEditors

0.33

 ungkapnya

0.33

Activations Density 0.027%

specific codes or identifiers likely related to software or programming contexts

hexadecimal strings and bytes

No Comments

No Known Activations