INDEX

Explanations

contextual references to being close or nearby

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Wikimedijinoj

-0.43

migrationBuilder

-0.42

 arquitecto

-0.42

 hombros

-0.41

 superiori

-0.41

 cejas

-0.40

 soeur

-0.40

 Menschheit

-0.40

pazität

-0.40

 Landschaft

-0.39

POSITIVE LOGITS

 proximity

0.87

Proximity

0.80

 Proximity

0.79

 closeness

0.77

closer

0.72

 closer

0.69

 Clos

0.69

 Closer

0.66

Closer

0.65

 closest

0.65

Activations Density 0.470%

contextual references to being close or nearby

No Comments

No Known Activations