INDEX

Explanations

references to validity in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 zelve

-0.47

 timpul

-0.47

 picioare

-0.47

 zijne

-0.47

 oreille

-0.46

 ziua

-0.45

 schoenen

-0.43

 zoude

-0.42

 conducción

-0.42

 argint

-0.42

POSITIVE LOGITS

valid

1.00

 valid

0.93

Valid

0.85

 nahilalakip

0.74

 Valid

0.69

 optional

0.65

sizeCache

0.65

VALID

0.63

 VALID

0.63

AddRange

0.57

Activations Density 0.328%

references to validity in various contexts

No Comments

No Known Activations