INDEX

Explanations

time-related references and updates

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

recent time periods

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

 ModelExpression

-0.60

 للمعارف

-0.59

Gambas

-0.52

ammans

-0.52

Tikang

-0.52

 specchio

-0.52

읗

-0.51

 sobą

-0.51

باردا

-0.49

makeConstraints

-0.49

POSITIVE LOGITS

 recently

0.66

Recently

0.61

 Recently

0.58

recently

0.50

 lately

0.50

最近

0.48

 recent

0.47

 recientemente

0.47

 недавно

0.46

 today

0.43

Activations Density 0.019%

time-related references and updates

recent time periods

No Comments

No Known Activations