INDEX

Explanations

key concepts related to evaluation and classification

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 ÎľÎ¿Î½

-0.15

erken

-0.15

strup

-0.15

nosti

-0.14

 fmap

-0.14

éĻ¸

-0.14

_inches

-0.13

apesh

-0.13

opers

-0.13

openh

-0.13

POSITIVE LOGITS

_sd

0.17

ening

0.17

enville

0.16

pie

0.15

iba

0.15

ahlen

0.15

arris

0.14

ikal

0.14

inz

0.14

ultz

0.14

Activations Density 0.017%

key concepts related to evaluation and classification

No Known Activations