INDEX

Explanations

phrases that indicate functions or roles of substances or materials in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 chofe

-0.68

 Monfieur

-0.63

 originaire

-0.60

 fhew

-0.58

 ſche

-0.57

 myſelf

-0.56

 pleaſure

-0.55

 avoient

-0.55

ChromeDriver

-0.54

 ſhe

-0.54

POSITIVE LOGITS

 dijadikan

0.75

 menjadi

0.72

 become

0.68

 became

0.67

 sebagai

0.64

作为

0.62

成为

0.61

 качестве

0.61

作為

0.60

 role

0.59

Activations Density 0.558%

phrases that indicate functions or roles of substances or materials in various contexts

No Comments

No Known Activations