INDEX

Explanations

higher values in context to mathematical or logical comparisons

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

greater than, bigger, upper

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

◄

-0.58

 ymin

-0.52

parsedMessage

-0.50

aarrggbb

-0.48

 referrerpolicy

-0.46

subpackage

-0.44

 CommonModule

-0.44

 lowest

-0.43

 lowering

-0.43

◀

-0.43

POSITIVE LOGITS

 positif

0.48

UVWXYZ

0.47

anyeol

0.47

posedge

0.43

bootstrapcdn

0.43

 positivo

0.42

LabelTagHelper

0.42

borderTop

0.42

 vermo

0.41

topRight

0.41

Activations Density 0.250%

higher values in context to mathematical or logical comparisons

greater than, bigger, upper

No Comments

No Known Activations