INDEX

Explanations

references to "elements" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

TintMode

-0.56

hit

-0.51

 injured

-0.50

 fine

-0.50

Tikang

-0.50

 pile

-0.49

Nya

-0.48

 BoxFit

-0.48

 defStyle

-0.48

interopRequire

-0.47

POSITIVE LOGITS

 elements

1.11

 element

1.08

Element

0.94

element

0.93

 elementos

0.91

 élément

0.90

 Elements

0.90

 Element

0.88

 Elemente

0.88

élément

0.87

Activations Density 0.318%

references to "elements" in various contexts

No Comments

No Known Activations