INDEX

Explanations

references to power dynamics and manipulation within hierarchical structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 يتيمه

-0.50

 onCreateView

-0.49

出版年

-0.42

Décès

-0.42

EndContext

-0.41

pyplot

-0.39

LabelTagHelper

-0.39

ANTLR

-0.38

rungsseite

-0.38

 Technical

-0.37

POSITIVE LOGITS

 assistants

0.67

 assistant

0.59

assistant

0.59

 Assistants

0.53

 fiel

0.53

 servants

0.52

 aides

0.51

loyal

0.50

 loyal

0.49

 uLocal

0.49

Activations Density 0.664%

references to power dynamics and manipulation within hierarchical structures

No Comments

No Known Activations