INDEX

Explanations

tokens and identifiers related to programming variables and functions, particularly in the context of coding structures and APIs

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

tz followed by dot, Shadow Split

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

EDEFAULT

-0.60

DoubleQuotes

-0.60

 defaultstate

-0.59

IsMutable

-0.58

RTSN

-0.57

 الحره

-0.56

 invokingState

-0.56

Viitteet

-0.52

 nakalista

-0.51

MigrationBuilder

-0.48

POSITIVE LOGITS

MethodManager

0.47

 يتيمه

0.40

webElementGuid

0.40

Controllo

0.38

 Subs

0.37

 Horst

0.36

évaluateur

0.36

 اكتوبر

0.35

Repeti

0.34

urch

0.34

Activations Density 0.587%

tokens and identifiers related to programming variables and functions, particularly in the context of coding structures and APIs

tz followed by dot, Shadow Split

No Comments

No Known Activations