INDEX

Explanations

non-English language characters, specifically from Asian scripts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

rians

-0.15

inan

-0.14

rian

-0.14

Atom

-0.14

ao

-0.14

abase

-0.14

ekler

-0.14

 Corm

-0.13

 Atom

-0.13

edu

-0.13

POSITIVE LOGITS

stdClass

0.15

à§į

0.15

enin

0.14

ASHBOARD

0.14

ð

0.14

RLF

0.13

 Hammer

0.13

éģ£

0.13

/runtime

0.13

ãĥ³ãĥĸ

0.13

Activations Density 0.025%

non-English language characters, specifically from Asian scripts

No Comments

No Known Activations