INDEX

Explanations

terms related to various types of advisors

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

seater

-0.47

GetAxis

-0.46

storey

-0.45

riceps

-0.45

Angry

-0.45

WebMethod

-0.44

ìm

-0.43

RenderAtEndOf

-0.43



-0.42

tolerant

-0.40

POSITIVE LOGITS

 exponential

0.62

 Township

0.59

 angel

0.59

 advisor

0.59

Endpoint

0.57

Advisor

0.56

Angel

0.56

 Appalachian

0.56

 Oscar

0.55

nm

0.55

Activations Density 0.252%

terms related to various types of advisors

No Comments

No Known Activations