INDEX

Explanations

words or phrases that imply a state of being affected or changed

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

OGND

-0.65

Autoritní

-0.62

 Roskov

-0.49

 Catawiki

-0.48

 cellspacing

-0.47

ACHUSET

-0.46

 iParam

-0.46

windowFixed

-0.44

obacter

-0.44

 itemprop

-0.43

POSITIVE LOGITS

pig

0.60

 pigment

0.59

 flush

0.59

Pig

0.58

Pig

0.53

pig

0.52

PIG

0.51

 drain

0.50

</thead>

0.49

 patrol

0.49

Activations Density 0.359%

words or phrases that imply a state of being affected or changed

No Comments

No Known Activations