INDEX

Explanations

transitional and coordinating phrases that signal relationships between ideas

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

conjunctions followed by clauses

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

RefNanny

-0.45

 hoor

-0.43

copyOf

-0.41

 PagesJaunes

-0.37

 صوتيه

-0.36

oredCriteria

-0.36

tangentMode

-0.36

styleType

-0.35

 rente

-0.35

 Meilleures

-0.34

POSITIVE LOGITS

rungsseite

0.51

outSlope

0.50

SequentialGroup

0.45

 Pingback

0.44

AccessorTable

0.44

skjaer

0.43

ISODE

0.43

apunov

0.42

CloseOperation

0.42

falgar

0.41

Activations Density 0.058%

transitional and coordinating phrases that signal relationships between ideas

conjunctions followed by clauses

No Comments

No Known Activations