INDEX

Explanations

references to the concept of tying, including different forms and contexts of the word "tie."

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

tie

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

 Aggarwal

-0.40

 unab

-0.39

Mr

-0.39

 magasiner

-0.38

creativecommons

-0.38

})).

-0.37

●●

-0.37

 Monat

-0.37

 mỗi

-0.36

NewGuid

-0.36

POSITIVE LOGITS

Tie

1.13

Tie

1.10

tie

1.09

 Ties

1.08

 ties

1.04

 tying

1.02

TIE

1.00

TIE

0.97

 tied

0.92

tie

0.90

Activations Density 0.007%

references to the concept of tying, including different forms and contexts of the word "tie."

tie

No Comments

No Known Activations