INDEX

Explanations

phrases related to instances or cases of events or occurrences

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

cases of instances

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

 banques

-0.42

 impuestos

-0.41

 plan

-0.41

konomi

-0.40

 Pave

-0.39

 PLAN

-0.38

labus

-0.38

attendre

-0.38

 clim

-0.38

 Reforms

-0.38

POSITIVE LOGITS

 instances

0.86

 occurrences

0.84

 occurrence

0.77

 cases

0.75

 Instances

0.73

 Fällen

0.70

occurrence

0.70

 incidents

0.69

 Fälle

0.68

 autorytatywna

0.67

Activations Density 0.039%

phrases related to instances or cases of events or occurrences

cases of instances

No Comments

No Known Activations