INDEX

Explanations

transition words and phrases that indicate contrast or addition in arguments

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

linking ideas

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

⟬

-0.86

expandindo

-0.84

ConstraintMaker

-0.75

LabelTagHelper

-0.72

 ModelExpression

-0.67

 كومونز

-0.63

ronpa

-0.60

 cherchés

-0.58

Hauptartikel

-0.56

 Audiodateien

-0.56

POSITIVE LOGITS

 kaarangay

0.34

chnet

0.31

ậc

0.30

下

0.28

bitset

0.28

पु

0.25

 PeEnEo

0.25

chauer

0.25

（

0.25

Activations Density 0.109%

transition words and phrases that indicate contrast or addition in arguments

linking ideas

No Comments

No Known Activations