INDEX

Explanations

phrases emphasizing assertions and evaluations related to quality or significance

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

quantifiers 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

 informée

-0.44

hoeddwyd

-0.35

 actionBar

-0.33

msgTypes

-0.32

canestro

-0.31

IBOutlet

-0.29

 Cases

-0.29

orianCalendar

-0.29

ViewById

-0.29

 مباشرة

-0.29

POSITIVE LOGITS

 defaultstate

0.60

ſelves

0.58

 vôtre

0.56

0.55

AsUp

0.54

 propOrder

0.53

 pleaſure

0.52

 reaſon

0.52

ükemmel

0.50

Activations Density 0.002%

phrases emphasizing assertions and evaluations related to quality or significance

quantifiers and measurements

No Comments

No Known Activations