INDEX

Explanations

temporal references and mentions of dates

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

years ago or dates

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

IntoConstraints

-0.75

parsedMessage

-0.68

DockStyle

-0.66

 snippetHide

-0.65

Diwedd

-0.65

 HasFactory

-0.65

 jsPsych

-0.62

Pens

-0.62

RTDA

-0.62

 EconPapers

-0.61

POSITIVE LOGITS

mişti

0.34

 habían

0.34

 había

0.33

 havia

0.33

 tinha

0.29

 aveva

0.28

 hadn

0.28

had

0.27

 habia

0.27

avait

0.27

Activations Density 0.169%

temporal references and mentions of dates

years ago or dates

No Comments

No Known Activations