INDEX

Explanations

references to various types of tracks and track-related activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ronics

-0.15

zed

-0.15

wen

-0.14

zeigen

-0.14

onet

-0.14

est

-0.14

cene

-0.13

kind

-0.13

onen

-0.13

POSITIVE LOGITS

side

0.15

yat

0.15

lobal

0.15

å·±

0.14

aden

0.14

iless

0.14

ieved

0.14

sumer

0.14

-muted

0.14

ÏĢÎŃ

0.14

Activations Density 0.017%

references to various types of tracks and track-related activities

No Comments

No Known Activations