INDEX

Explanations

various forms of the word "pattern" and related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/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.11.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

-0.54

 Wohnzimmer

-0.53

issauga

-0.48

 nahilalakip

-0.48

 Kühlschrank

-0.47

 Internasional

-0.47

 jopa

-0.45

:]:

-0.45

 sepenuhnya

-0.43

⁄

-0.42

POSITIVE LOGITS

 pattern

1.98

 Pattern

1.88

pattern

1.85

 PATTERN

1.78

 patterns

1.77

Pattern

1.75

 Patterns

1.68

PATTERN

1.55

patterns

1.49

Patterns

1.48

Activations Density 0.172%

various forms of the word "pattern" and related concepts

No Comments

No Known Activations