INDEX

Explanations

references to ongoing or future actions related to decision-making

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

Continue reading links

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

Aholisi

-0.34

RefNanny

-0.34

("}");

-0.29

se

-0.29

saco

-0.28

 onCancelled

-0.28

 EClass

-0.28

SourceChecksum

-0.28

 truy

-0.28

 collections

-0.28

POSITIVE LOGITS

 estekak

0.67

 news

0.63

 notizie

0.60

 AssemblyTitle

0.60

dataclass

0.58

 للاسماء

0.58

 berita

0.58

 informée

0.57

 queſto

0.57

 nieuws

0.57

Activations Density 0.021%

references to ongoing or future actions related to decision-making

Continue reading links

No Comments

No Known Activations