INDEX

Explanations

technical terms related to programming or software services

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

code files and paths

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

ConverterFactory

-0.40

mtrl

-0.40

setVerticalGroup

-0.39

 autorytatywna

-0.39

rungsseite

-0.38

enterOuterAlt

-0.36

 ${({

-0.35

 Chwiliwch

-0.34

yntaxException

-0.34

 <=",

-0.33

POSITIVE LOGITS

 nawr

0.60

 NSCoder

0.58

 capitolo

0.53

weightedMode

0.52

sifs

0.52

 nôtre

0.50

󠁢

0.50

itd

0.50

mnop

0.49

hope

0.49

Activations Density 0.091%

technical terms related to programming or software services

code files and paths

No Comments

No Known Activations