INDEX

Explanations

references to specific models or constructs in technical contexts

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

Followed by "was" or "were."

np_acts-logits-general · gemini-2.0-flash

statement was

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lampa

-0.41

/**

-0.41

AutoModerator

-0.37

 AppComponent

-0.37

ANTLR

-0.36

fileID

-0.36

 SqlDbType

-0.35

Jereo

-0.35

 calculées

-0.34

inola

-0.34

POSITIVE LOGITS

culable

0.47

ondissement

0.45

 ErrIntOverflow

0.42

enumi

0.38

 Wartawan

0.37

sune

0.37

 Vesu

0.37

 cassa

0.37

Jîn

0.37

 հղումներ

0.37

Activations Density 0.990%

references to specific models or constructs in technical contexts

Followed by "was" or "were."

statement was

No Comments

No Known Activations