INDEX

Explanations

references to extended durations of time or lengthy processes

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

long duration

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

ništvo

-0.48

 consciência

-0.47

ânico

-0.47

grà

-0.47

 gouvern

-0.46

 espírito

-0.46

 coscienza

-0.45

apollo

-0.45

JNIEnv

-0.45

 jiwa

-0.44

POSITIVE LOGITS

 lengthy

0.94

 long

0.90

long

0.71

 langen

0.70

 LONG

0.70

 lange

0.69

長い

0.69

 Long

0.68

 longue

0.68

 dài

0.68

Activations Density 0.022%

references to extended durations of time or lengthy processes

long duration

No Comments

No Known Activations