INDEX

Explanations

contextual and relational phrases that indicate feelings or opinions about situations

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

Followed by "so" (therefore/thus/accordingly)

np_acts-logits-general · gemini-2.0-flash

consequence or conclusion

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.16.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

رشف

-0.68

 Meksiku

-0.64

Efq

-0.61

bouncycastle

-0.58

клопе

-0.57

otherwise

-0.56

httphttps

-0.56

Otherwise

-0.55

енча

-0.55

ſelves

-0.54

POSITIVE LOGITS

จึง

0.80

 derfor

0.76

わけです

0.72

 accordingly

0.71

 więc

0.69

 donc

0.69

sprechend

0.65

也就

0.64

 därför

0.64

 Therefore

0.63

Activations Density 0.470%

contextual and relational phrases that indicate feelings or opinions about situations

Followed by "so" (therefore/thus/accordingly)

consequence or conclusion

No Comments

No Known Activations