INDEX

Explanations

quotes or phrases that indicate someone is speaking or expressing an opinion

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

attributing speech

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

الحياه

-0.73

:✨

-0.68

 kasarigan

-0.67

 autorytatywna

-0.65

Aholisi

-0.64

bootstrapcdn

-0.63

 FetchType

-0.59

 Мексичка

-0.59

تقاوى

-0.59

tagHelperRunner

-0.55

POSITIVE LOGITS

0.59

himself

0.47

 during

0.47

 himself

0.45

 when

0.43

 ujarnya

0.43

 said

0.42

during

0.42

。

0.42

 aldus

0.41

Activations Density 0.009%

quotes or phrases that indicate someone is speaking or expressing an opinion

attributing speech

No Comments

No Known Activations