INDEX

Explanations

terms and concepts associated with definitions and known entities

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

Preceding words of "as"

np_acts-logits-general · gemini-2.0-flash

defining or calling "as"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 disambiguazione

-0.78

تقاوى

-0.77

 חיצוניים

-0.70

findpost

-0.70

 AssemblyCompany

-0.69

 وتسجيلات

-0.65

enumi

-0.63

IUrlHelper

-0.62

AndEndTag

-0.62

enderror

-0.62

POSITIVE LOGITS

as

3.06

 sebagai

2.35

 jako

1.92

 ως

1.82

作为

1.82

作為

1.70

sebagai

1.60

 Sebagai

1.60

 как

1.58

 као

1.57

Activations Density 2.207%

terms and concepts associated with definitions and known entities

Preceding words of "as"

defining or calling "as"

No Comments

No Known Activations