INDEX

Explanations

references to collective experiences and shared involvement

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

we, us, and our

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

 iprot

-0.36

"}")

-0.36

 Vereine

-0.35

 kasarigan

-0.35



-0.34

 Numerade

-0.34

ёв

-0.33

LookAnd

-0.33

الدراسه

-0.32

 actionMode

-0.32

POSITIVE LOGITS

principalTable

0.52

schild

0.47

 digested

0.46

GOTREF

0.45

 ſind

0.43

 anhydride

0.43

respect

0.42

耘

0.42

⤹

0.42

 team

0.42

Activations Density 0.025%

references to collective experiences and shared involvement

we, us, and our

No Comments

No Known Activations