INDEX

Explanations

references to academic or research-related websites

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

èĬĿ

-0.15

illet

-0.14

ODE

-0.14

Äįer

-0.14

otron

-0.14

aeda

-0.14

 Å¡ÃŃ

-0.14

arrow

-0.14

ÏĦÎ¬

-0.14

okit

-0.13

POSITIVE LOGITS

.synthetic

0.18

ellido

0.16

nila

0.15

 noreferrer

0.14

lando

0.14

uhe

0.14

 absor

0.13

aptive

0.13

uptools

0.13

scri

0.13

Activations Density 0.005%

references to academic or research-related websites

No Comments

No Known Activations