INDEX

Explanations

phrases indicating the highest quality or top rankings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

rl

-0.15

onic

-0.14

icina

-0.14

 Distrib

-0.14

mp

-0.14

ry

-0.14

eri

-0.14

rs

-0.14

ble

-0.14

ulla

-0.14

POSITIVE LOGITS

ardo

0.15

.ret

0.14

_inches

0.14

lashes

0.14

iaz

0.14

ret

0.14

dued

0.14

 å§

0.14

.utf

0.14

angs

0.14

Activations Density 0.015%

phrases indicating the highest quality or top rankings

No Comments

No Known Activations