INDEX

Explanations

instances of the word "which" and its variations to identify clauses or additional information in sentences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

leſs

-0.52

 Landschaft

-0.48

ster

-0.47

 kleid

-0.47

odyear

-0.46

nate

-0.45

 draußen

-0.44

anera

-0.44

ptid

-0.43

lossene

-0.43

POSITIVE LOGITS

 WHICH

1.11

 Which

1.06

Which

0.98

which

0.91

 which

0.91

 wich

0.83

hich

0.78

 Wich

0.76

 والتي

0.71

Wich

0.67

Activations Density 0.213%

instances of the word "which" and its variations to identify clauses or additional information in sentences

No Comments

No Known Activations