INDEX

Explanations

references to the word 'gan' and its variations, along with geometrical terms related to triangles

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Cyfeiriadau

-0.52

脚注の使い方

-0.50

 colorear

-0.43

 intStringLen

-0.40

ñador

-0.38

кобрита

-0.37

一听

-0.36

Lähteet

-0.36

 Página

-0.35

uxxxx

-0.35

POSITIVE LOGITS

gan

0.61

 triangle

0.60

 pole

0.57

unfinished

0.56

 lottery

0.54

 fixtures

0.53

 dynasties

0.52

 triangles

0.52

Ö

0.51

ö

0.51

Activations Density 0.565%

references to the word 'gan' and its variations, along with geometrical terms related to triangles

No Known Activations