INDEX

Explanations

occurrences of the word "which" in various forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

-0.59

enal

-0.51

 Kalis

-0.50

odyear

-0.49

 Landschaft

-0.48

ptid

-0.47

odle

-0.46

ster

-0.46

anera

-0.46

 kleid

-0.46

POSITIVE LOGITS

 WHICH

0.97

 Which

0.94

Which

0.86

which

0.85

 which

0.83

 wich

0.78

 Wich

0.71

Wich

0.69

hich

0.69

 والتي

0.68

Activations Density 0.225%

occurrences of the word "which" in various forms

No Comments

No Known Activations