INDEX

Explanations

references to engine performance under varying load conditions

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

 Paglinawan

-0.40

 cầm

-0.38

 unknownFields

-0.37

zab

-0.36

amarca

-0.35

!("{}",

-0.35

 Dominican

-0.34

Jut

-0.34

iseite

-0.33

 Vikipedi

-0.33

POSITIVE LOGITS

 load

3.63

 Load

3.17

load

3.09

Load

2.92

 LOAD

2.91

 loading

2.84

 loads

2.83

 loaded

2.63

LOAD

2.59

 Loads

2.58

Activations Density 0.762%

references to engine performance under varying load conditions

No Comments

No Known Activations