INDEX

Explanations

the word "recently" as an indicator of recent events or occurrences

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

recent past events

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

 propOrder

-0.45

 pleaſure

-0.41

 BoxFit

-0.41

PDG

-0.41

 Kohler

-0.40

 maș

-0.40

 chio

-0.40

hme

-0.40

MQ

-0.39

 iscri

-0.39

POSITIVE LOGITS

recently

1.12

 recently

1.03

 Recently

0.94

Recently

0.92

 recientemente

0.84

recent

0.82

Recent

0.76

 recentemente

0.75

 RECENT

0.75

 lately

0.73

Activations Density 0.019%

the word "recently" as an indicator of recent events or occurrences

recent past events

No Comments

No Known Activations