INDEX

Explanations

scientific measurements and statistical analysis related to experimental data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.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

requent

-0.36

 feroit

-0.35

 přid

-0.35

hdys

-0.34

heitsbild

-0.34

 bizony

-0.34

今から

-0.34

 souhait

-0.33

moveToFirst

-0.33

 kepentingan

-0.33

POSITIVE LOGITS

 total

0.92

 average

0.91

 Total

0.75

 ratio

0.73

 overall

0.72

 mean

0.72

 Average

0.72

平均

0.68

total

0.67

Average

0.67

Activations Density 1.919%

scientific measurements and statistical analysis related to experimental data

No Comments

No Known Activations