INDEX

Explanations

code-related structures and functions pertaining to data manipulation, particularly involving lists and nodes

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

Code or data followed by specific keywords

np_acts-logits-general · gemini-2.0-flash

current phase value

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

 kasarigan

-0.55

DoubleQuotes

-0.54



-0.54

 skies

-0.51

 Wohnen

-0.51

 fats

-0.50

 lows

-0.49

 grasses

-0.49

 gangs

-0.48

nesty

-0.48

POSITIVE LOGITS

+#+#

0.72

Izvori

0.67

=$?

0.67

 noDo

0.66

ſtanding

0.63

 himſelf

0.61

 themſelves

0.61

州市

0.60

 Theſe

0.59

gameserver

0.57

Activations Density 0.357%

code-related structures and functions pertaining to data manipulation, particularly involving lists and nodes

Code or data followed by specific keywords

current phase value

No Comments

No Known Activations