INDEX

Explanations

words related to time, events, and their associated contexts.

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

dates, times and measurements

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

 initComponents

-0.60

 Económica

-0.49

 perfección

-0.44

IBOutlet

-0.44

libatkan

-0.43

 surla

-0.42

 personnalité

-0.41

IndentedString

-0.41

retudo

-0.41

 cejas

-0.40

POSITIVE LOGITS

 OMITBAD

0.62

RTEE

0.62

rungsseite

0.56

存于互联网档案馆

0.52

AsUp

0.51

emlrt

0.49

 insuffisamment

0.48

冠

0.48

HttpFoundation

0.45

+#+#

0.45

Activations Density 0.377%

words related to time, events, and their associated contexts.

dates, times and measurements

No Comments

No Known Activations