INDEX

Explanations

high-frequency words and phrases indicative of questions and arguments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

opsis

-0.16

 Strawberry

-0.15

vÄĽd

-0.15

CAPE

-0.15

esti

-0.14

iforn

-0.14

 prim

-0.14

HashCode

-0.14

POSITIVE LOGITS

umo

0.16

irst

0.16

.fm

0.15

gable

0.15

IRST

0.14

vem

0.14

à¹ģà¸Ļ

0.14

EEK

0.14

 Nicholson

0.14

ogg

0.14

Activations Density 0.008%

high-frequency words and phrases indicative of questions and arguments

No Comments

No Known Activations