INDEX

Explanations

conjunctions and the relationships between items in a list

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

lists of specific nouns

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

Encore

-0.34

ropoda

-0.33

literal

-0.32

 kokona

-0.31

fiction

-0.31

 côtés

-0.31

 inkluder

-0.31

ufnahme

-0.30

leit

-0.30

 samfun

-0.30

POSITIVE LOGITS

.*")]

0.69

RenderAtEndOf

0.60

QMetaType

0.58

msgTypes

0.55

 Photocase

0.54

 AssemblyTitle

0.53

 InputDecoration

0.53

 فريبيس

0.53

 tartalomajánló

0.52

 незавершена

0.52

Activations Density 0.203%

conjunctions and the relationships between items in a list

lists of specific nouns

No Comments

No Known Activations