INDEX

Explanations

references to a context variable in code

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

object variable prefixes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Slf

-0.66

Portail

-0.63

новниш

-0.61

 مشين

-0.58

getWriter

-0.56

tvguidetime

-0.54

 wind

-0.54

bufio

-0.53

Искәрмәләр

-0.52

Portály

-0.52

POSITIVE LOGITS

 mContext

0.91

mContext

0.73

 mCurrent

0.64

 mView

0.60

 getItemCount

0.59

 mList

0.56

 AssemblyTitle

0.54

 mAdapter

0.53

 mData

0.49

mTitle

0.47

Activations Density 0.008%

references to a context variable in code

object variable prefixes

No Comments

No Known Activations