INDEX

Explanations

references to numerical values and comparisons within the context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

<bos>

-0.57

 indeed

-0.41

to

-0.39

 including

-0.37

 impenetrable

-0.36

 تص

-0.36

 third

-0.36

 unintended

-0.36

 MacLeod

-0.36

POSITIVE LOGITS

 values

2.25

 Values

2.14

values

2.05

Values

2.02

 VALUES

1.91

VALUES

1.77

 valores

1.70

 Werte

1.63

 valeurs

1.63

 Valores

1.48

Activations Density 0.174%

references to numerical values and comparisons within the context

No Comments

No Known Activations