INDEX

Explanations

instances of the letter 'H' in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

elong

-0.17

uper

-0.17

andas

-0.16

utenberg

-0.16

ids

-0.16

HY

-0.15

reset

-0.15

incare

-0.15

ãģĤãģĴ

-0.15

oci

-0.15

POSITIVE LOGITS

IGHL

0.27

OSP

0.27

ISP

0.24

OLLOW

0.24

ILLS

0.24

IGHLIGHT

0.24

OMEM

0.22

ORIZONTAL

0.21

ERSHEY

0.21

OLID

0.21

Activations Density 0.010%

instances of the letter 'H' in various contexts

No Comments

No Known Activations