INDEX

Explanations

transitional phrases that indicate a shift in topic or direction

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

returning to a topic

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

期刊论文

-0.41

 jeweils

-0.38

ulgação

-0.37

ReadByte

-0.36

 saling

-0.36

 Dabei

-0.35

ThemeOverlay

-0.35

chafter

-0.35

ratore

-0.34

TagMode

-0.34

POSITIVE LOGITS

 digress

0.94

 topic

0.75

に戻

0.73

Zurück

0.72

 tillbaka

0.70

回到

0.68

 zurück

0.67

Returning

0.67

 kembali

0.65

 Topic

0.65

Activations Density 0.040%

transitional phrases that indicate a shift in topic or direction

returning to a topic

No Known Activations