INDEX

Explanations

expressions related to effort and responsibility in actions

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

actions followed by intent

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

 finally

-0.51

 наконец

-0.49

 finalmente

-0.48

었습니다

-0.47

ktı

-0.45

was

-0.44

 Finally

-0.43

 prevailed

-0.43

 recently

-0.43

 prematurely

-0.43

POSITIVE LOGITS

 Савезне

0.59

Fordítás

0.55

 regularly

0.54

 actualidad

0.54

 Biôgrafia

0.53

regular

0.52

rungsseite

0.50

 AssemblyCompany

0.49

 surla

0.49

 चीज़ों

0.47

Activations Density 0.271%

expressions related to effort and responsibility in actions

actions followed by intent

No Comments

No Known Activations