INDEX

Explanations

terms related to the concept of "recent" or occurrences within a recent timeframe

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

recent or coming time units

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

⟬

-0.69

rungsseite

-0.65

 propOrder

-0.63

gonic

-0.60

Datuak

-0.60

 kasarigan

-0.59

 Aufenthalt

-0.59

ніципалі

-0.58

findpost

-0.57



-0.57

POSITIVE LOGITS

 years

0.56

 times

0.54

few

0.50

 tqdm

0.49

 days

0.49

 decades

0.46

 OkHttpClient

0.41

 weeks

0.40

 endforeach

0.40

 past

0.40

Activations Density 0.009%

terms related to the concept of "recent" or occurrences within a recent timeframe

recent or coming time units

No Comments

No Known Activations