INDEX

Explanations

instances of the word "think" and its variations, indicating a focus on thoughts and reflection

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

 Guill

-0.50

 Maurer

-0.46

 Jacobsen

-0.46

 Joel

-0.46

 Jaffe

-0.44

 הס

-0.42

onerror

-0.42

»

-0.42

«

-0.42

 Quig

-0.42

POSITIVE LOGITS

 Think

1.44

Think

1.41

 THINK

1.41

 think

1.40

think

1.36

THINK

1.29

 thinks

1.26

 thought

1.16

 Thinking

1.15

thought

1.13

Activations Density 0.124%

instances of the word "think" and its variations, indicating a focus on thoughts and reflection

No Comments

No Known Activations