INDEX

Explanations

terms related to programming libraries and code structures

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

Code diff or version control markers

np_acts-logits-general · gemini-2.0-flash

Text variations

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

 surla

-0.56

 يتيمه

-0.54

polation

-0.49

 محفوظة

-0.48

 hatır

-0.47

jsonwebtoken

-0.45

 كومونز

-0.45

 مرئيه

-0.45

ांकि

-0.44

RenderAtEndOf

-0.44

POSITIVE LOGITS

Text

1.07

 Text

0.81

 text

0.68

text

0.64

WarningLevel

0.56

adaptiveStyles

0.55

 Texte

0.54

TEXT

0.54

 texts

0.52

 texte

0.51

Activations Density 0.009%

terms related to programming libraries and code structures

Code diff or version control markers

Text variations

No Comments

No Known Activations