INDEX

Explanations

references to viewing articles and sources in a scholarly context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pez

-0.16

æĶĿ

-0.15

OwnProperty

-0.15

DFS

-0.14

onec

-0.14

 undert

-0.14

/loose

-0.14

Sad

-0.14

POSITIVE LOGITS

áº¿t

0.19

UBE

0.16

ube

0.16

Î»Î®ÏĤ

0.15

 onHide

0.15

ÅĤad

0.15

zin

0.14

à¸Ĳà¸²à¸Ļ

0.14

ÙĪÙĩ

0.14

 horizon

0.14

Activations Density 0.085%

references to viewing articles and sources in a scholarly context

No Comments

No Known Activations