INDEX

Explanations

conjunctions and terms expressing comparisons or similarities

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

"and", "too," or similar conjunctions

np_acts-logits-general · gemini-2.0-flash

conjunctions, connector words

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

stdc

-0.85

 ModelRenderer

-0.82

@"/

-0.73

 Ceinture

-0.72

featureID

-0.71

 betweenstory

-0.69

Personensuche

-0.66

 newBuilder

-0.66

TypedDataSet

-0.65

 surla

-0.65

POSITIVE LOGITS

たまた

0.85

 esterni

0.71

 kasarigan

0.64

즉

0.59

 zwar

0.57

</i>

0.56

fortawesome

0.56

 חיצוניים

0.52

ouille

0.51

Koordinaten

0.51

Activations Density 1.437%

conjunctions and terms expressing comparisons or similarities

"and", "too," or similar conjunctions

conjunctions, connector words

No Comments

No Known Activations