INDEX

Explanations

references to research and academic work

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

owie

-0.16

Î³Î³

-0.16

ows

-0.15

ays

-0.14

eph

-0.14

è¾º

-0.14

loo

-0.14

inx

-0.14

czy

-0.14

POSITIVE LOGITS

ifi

0.25

icer

0.23

iten

0.22

ifer

0.21

IFI

0.20

icon

0.20

ius

0.20

isc

0.19

ient

0.19

iforn

0.19

Activations Density 0.003%

references to research and academic work

No Comments

No Known Activations

references to research and academic work

No Comments

No Known Activations