INDEX

Explanations

references to computer code or technical programming elements

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

"package"

np_acts-logits-general · gemini-2.0-flash

package id

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 enfans

-0.61

 prestazioni

-0.56

 piemē

-0.55

 distanciation

-0.55

Amazing

-0.54

 vanske

-0.53

principalTable

-0.52

 tjen

-0.52

 intptr

-0.51

lapsingToolbar

-0.49

POSITIVE LOGITS

<bos>

0.94

 createSlice

0.74

=$?

0.66

ynos

0.57

com

0.57

]?

0.56

Bronnen

0.55

package

0.55

)?;

0.55

 مشارکت‌کنندگان

0.54

Activations Density 0.057%

references to computer code or technical programming elements

"package"

package id

No Comments

No Known Activations