INDEX

Explanations

numerical identifiers related to specific items or editions

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

ospels

-0.16

ÑĢÐ¸Ñĩ

-0.15

reh

-0.14

uali

-0.14

lez

-0.14

ypad

-0.14

isel

-0.14

.Keyword

-0.14

;line

-0.14

gend

-0.13

POSITIVE LOGITS

0.17

0.15

 Maher

0.14

sted

0.14

 Temper

0.14

0.13

 Leone

0.13

atica

0.13

Activations Density 0.001%

numerical identifiers related to specific items or editions

No Comments

No Known Activations