INDEX

Explanations

positive descriptions of locations or options

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 skóra

-0.52

:+:

-0.46

 lumea

-0.42

Rujuakan

-0.42

 OFDb

-0.41

 resourceCulture

-0.41

ValueStyle

-0.40

 ovlád

-0.40

 černá

-0.39

InSection

-0.39

POSITIVE LOGITS

 hotels

0.48

ressee

0.46

 مشين

0.46

 Recommend

0.45

InputBorder

0.45

最快更新

0.44

 Preference

0.44

RECOMMENDATIONS

0.44

recommended

0.44

 recommend

0.44

Activations Density 0.001%

positive descriptions of locations or options

No Comments

No Known Activations