INDEX

Explanations

titles or names of TV shows and movies

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

committed

-0.15

ault

-0.14

aminer

-0.14

.skin

-0.14

vere

-0.13

uele

-0.13

ãĥģ

-0.13

 Voor

-0.13

Analytics

-0.13

uner

-0.12

POSITIVE LOGITS

ibraries

0.17

PLL

0.15

udder

0.15

pillar

0.15

æ··

0.15

ÐµÐ½Ñģ

0.14

InternalServerError

0.14

pike

0.13

 redirects

0.13

istrovstvÃŃ

0.13

Activations Density 0.118%

titles or names of TV shows and movies

No Comments

No Known Activations