INDEX

Explanations

sections that provide instructions or commands

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

illo

-0.19

aber

-0.15

ives

-0.15

ÐµÐ»Ñİ

-0.14

echo

-0.14

Dot

-0.13

ppt

-0.13

cker

-0.13

otts

-0.13

ar

-0.13

POSITIVE LOGITS

ieee

0.16

bourg

0.15

stery

0.14

erdem

0.14

erne

0.14

ermen

0.14

yb

0.13

inyin

0.13

faf

0.13

.gnu

0.13

Activations Density 0.007%

sections that provide instructions or commands

No Comments

No Known Activations