INDEX

Explanations

terms related to computational techniques and their effectiveness in data processing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 imprimible

-0.56

 AssemblyCompany

-0.54

 navideño

-0.52

-------------</

-0.49

 inoxidable

-0.47

chufe

-0.47

SourceChecksum

-0.47

luents

-0.45

เหร

-0.44

 provisoire

-0.44

POSITIVE LOGITS

 recomm

0.67

 recommendation

0.67

 recommend

0.59

 Recom

0.58

 Recommendation

0.57

Recom

0.56

 recommending

0.55

 Recommend

0.55

 similarity

0.55

 relevance

0.54

Activations Density 0.750%

terms related to computational techniques and their effectiveness in data processing

No Comments

No Known Activations