INDEX

Explanations

numerical values, particularly those that signify importance or rank

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

Code, filenames, or programming-related text

np_acts-logits-general · gemini-2.0-flash

introduces phrases or clauses

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

 ejus

-0.54

Vidite

-0.52

resave

-0.51

ymal

-0.50

 tige

-0.50

 JAXBElement

-0.49

MatIconModule

-0.49

GEBURTS

-0.48

 IconButton

-0.48

 eorum

-0.47

POSITIVE LOGITS

 Audiodateien

0.73

 мәкал

0.62

 nakalista

0.61

ContentAsync

0.59

 AssemblyVersion

0.57

 الحره

0.56

hode

0.53

Hochspringen

0.52

stdc

0.51

initComponents

0.50

Activations Density 0.033%

numerical values, particularly those that signify importance or rank

Code, filenames, or programming-related text

introduces phrases or clauses

No Comments

No Known Activations