INDEX

Explanations

words that denote the preparation or acquisition of materials or data within scientific contexts

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

obtained results and materials

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

 auroit

-0.68

Portály

-0.65

aarrggbb

-0.64

 pouvoit

-0.61

 ainfi

-0.61

 ſon

-0.61

 Monfieur

-0.60

 feroit

-0.59

Portail

-0.58

了嗎

-0.57

POSITIVE LOGITS

 Used

0.38

(&:

0.36

 currents

0.34

ActionPerformed

0.34

Available

0.33

 measured

0.33

Used

0.33

 medications

0.33

 pist

0.33

 ongoing

0.33

Activations Density 0.096%

words that denote the preparation or acquisition of materials or data within scientific contexts

obtained results and materials

No Comments

No Known Activations