INDEX

Explanations

references to established or recognized information, issues, or facts

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

iphy

-0.16

gli

-0.16

 nett

-0.16

 blanks

-0.15

ars

-0.15

ross

-0.14

 Blasio

-0.14

gro

-0.14

Ð¾ÑģÑĤ

-0.14

infeld

-0.14

POSITIVE LOGITS

ÙĨØª

0.18

ledge

0.16

ienes

0.15

ÑģÑĮ

0.15

rops

0.14

Ø©

0.14

aghan

0.14

ÎµÏĢ

0.14

orte

0.14

enus

0.14

Activations Density 0.037%

references to established or recognized information, issues, or facts

No Comments

No Known Activations