INDEX

Explanations

references to varying levels of concern or importance

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

Follows "great," "due," or "considerable."

np_acts-logits-general · gemini-2.0-flash

great positive descriptors

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

AllAfrica

-0.64

AccessException

-0.60

antMatchers

-0.60

 ApiException

-0.58

 unconditionally

-0.57

Geplaatst

-0.56

aarrggbb

-0.56

 poveznice

-0.55

 esercit

-0.54

 egens

-0.53

POSITIVE LOGITS

Hochspringen

0.56

queryInterface

0.56

 concern

0.55

<bos>

0.55

wenn

0.54

schild

0.51

KEYCODE

0.50

 scrutiny

0.48

govina

0.47

EDEFAULT

0.46

Activations Density 0.574%

references to varying levels of concern or importance

Follows "great," "due," or "considerable."

great positive descriptors

No Comments

No Known Activations