INDEX

Explanations

instances of the word "be."

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

can be + past participle

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

io

-0.40

-0.39

 Ster

-0.37

-0.35

 rangs

-0.35

↵↵

-0.34

 pouvoir

-0.34

 Ziegler

-0.34

POSITIVE LOGITS

хьтан

0.93

 canst

0.84

easily

0.84

CanBe

0.82

 easily

0.81

<unused52>

0.79

<unused68>

0.79

<unused41>

0.79

[@BOS@]

0.79

<unused28>

0.79

Activations Density 0.080%

instances of the word "be."

can be + past participle

No Comments

No Known Activations