INDEX

Explanations

references to companies and firms in various contexts

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

companies and their attributes

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

 gemak

-0.47

 occurring

-0.40

 szczegól

-0.37

 olacaktır

-0.36

 głę

-0.35

 towarzys

-0.35

 geleg

-0.35

 deberes

-0.35

 reflect

-0.35

 depth

-0.35

POSITIVE LOGITS

 company

1.11

 companies

1.11

 Companies

0.98

 empresa

0.97

 Company

0.92

Companies

0.92

 entreprise

0.90

companies

0.90

 perusahaan

0.89

 AssemblyCompany

0.88

Activations Density 0.040%

references to companies and firms in various contexts

companies and their attributes

No Comments

No Known Activations