INDEX

Explanations

occurrences of the word "connect" and its variations in the context of networking or connectivity

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

connect followed by punctuation or word

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ia

-0.41

ian

-0.39

 usual

-0.35

(/\

-0.34

––––

-0.33

aria

-0.33

Ley

-0.33

hood

-0.33

 Major

-0.32

POSITIVE LOGITS

Connect

1.93

 Connect

1.92

connect

1.62

CONNECT

1.58

 CONNECT

1.51

 connect

1.18

onnect

1.14

 Conne

1.06

Conne

1.05

connects

1.03

Activations Density 0.012%

occurrences of the word "connect" and its variations in the context of networking or connectivity

connect followed by punctuation or word

No Comments

No Known Activations