INDEX

Explanations

numeric identifiers and references related to various categories or types of products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

eturn

-0.17

robat

-0.17

ayet

-0.14

mey

-0.14

orsk

-0.14

oler

-0.14

uge

-0.14

olk

-0.14

isu

-0.14

vinc

-0.14

POSITIVE LOGITS

?url

0.16

.gf

0.15

 BÃ¶l

0.14

AAD

0.14

Ø³ØªØ§ÙĨ

0.14

reopen

0.14

#ac

0.14

subclass

0.14

 Aspen

0.13

igham

0.13

Activations Density 0.789%

numeric identifiers and references related to various categories or types of products

No Comments

No Known Activations