INDEX

Explanations

phrases and numerical indicators related to quantity or distance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

 entire

-0.56

 ENTIRE

-0.49

entire

-0.47

DockStyle

-0.47

SequentialGroup

-0.45

longer

-0.44

 MINIMUM

-0.44

minimum

-0.43

 WHOLE

-0.43

 Entire

-0.42

POSITIVE LOGITS

 doors

0.60

 fhort

0.60

 stones

0.59

0.58

 short

0.55

MemoryWarning

0.54

 steps

0.53

Captor

0.51

 strides

0.50

 paces

0.50

Activations Density 0.225%

phrases and numerical indicators related to quantity or distance

No Comments

No Known Activations