INDEX

Explanations

references to planning and scheduling events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.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

 سكانية

-0.49

 hereof

-0.41

 Evidently

-0.40

 kasarigan

-0.40

 thereof

-0.38

stylers

-0.36

domada

-0.36

 apprehensions

-0.35

 logarithms

-0.35

 例证

-0.35

POSITIVE LOGITS

 dunno

0.63

 kena

0.58

 memang

0.57

 cannot

0.56

dun

0.56

 bodo

0.55

 macam

0.55

oso

0.55

 lubang

0.54

ppl

0.53

Activations Density 0.300%

references to planning and scheduling events

No Comments

No Known Activations