INDEX

Explanations

expressions or conjunctions that indicate relationships between characters or elements in a narrative

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

wearing clothes and accessories

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

 partnership

-0.35

 economic

-0.34

 community

-0.34

 region

-0.33

 development

-0.33

olescence

-0.33

 expansion

-0.32

 PeEnEo

-0.31

 transférez

-0.31

 developed

-0.31

POSITIVE LOGITS

 jspb

0.63

 BoxFit

0.56

 smelling

0.55

 grinning

0.54

MethodManager

0.53

 Audiodateien

0.52

AddHtmlAttribute

0.51

 mustache

0.50

wearing

0.50

Carriera

0.50

Activations Density 0.033%

expressions or conjunctions that indicate relationships between characters or elements in a narrative

wearing clothes and accessories

No Comments

No Known Activations