INDEX

Explanations

punctuations and conjunctions that indicate conditions or attributes in contexts

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

strict not strict

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

 виправивши

-0.91

GEBURTS

-0.80

 resourceCulture

-0.73

+#+#

-0.68

Hentet

-0.68

+#+

-0.67

bootstrapcdn

-0.66

 BoxFit

-0.64

 صوتيه

-0.63

 فريبيس

-0.60

POSITIVE LOGITS

("")]

0.47

 duradero

0.43

 câmbio

0.43

 vêtement

0.43

équipements

0.43

 oamen

0.43

InjectAttribute

0.43

 praktijk

0.42

 pihaknya

0.42

 Bewußt

0.42

Activations Density 0.261%

punctuations and conjunctions that indicate conditions or attributes in contexts

strict not strict

No Comments

No Known Activations