INDEX

Explanations

various references to properties and methods in programming contexts, particularly related to session management and data handling

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

Appears before words in other languages

np_acts-logits-general · gemini-2.0-flash

multilingual instruction following

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

 Theſe

-1.59

 myſelf

-1.58

Efq

-1.55

 Monfieur

-1.52

 itſelf

-1.47

 Shakspeare

-1.44

 Jefus

-1.40

 raiſ

-1.40

 ſeveral

-1.40

 fubject

-1.37

POSITIVE LOGITS

↵

0.56

<eos>

0.54

0.47

↵↵

0.45

</td>

0.43

<unused63>

0.42

↵↵↵

0.42

</h3>

0.42

<unused61>

0.41

Activations Density 0.055%

various references to properties and methods in programming contexts, particularly related to session management and data handling

Appears before words in other languages

multilingual instruction following

No Comments

No Known Activations