INDEX

Explanations

modal verbs indicating potential future actions or outcomes

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

ability and possibility

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

 للمعارف

-0.54

Tembelea

-0.50

thenReturn

-0.44

msford

-0.44

RTEE

-0.44

 beginnetje

-0.43

expandindo

-0.43

superman

-0.41

 hypot

-0.40

 HFILL

-0.40

POSITIVE LOGITS

 <>",

0.45

.*")]

0.42

 voorbere

0.41

 pido

0.40

 sienta

0.40

 "..\..\..\

0.38

 Wicidata

0.38

bezpe

0.38

 acuer

0.35

 voglio

0.35

Activations Density 0.096%

modal verbs indicating potential future actions or outcomes

ability and possibility

No Comments

No Known Activations