INDEX

Explanations

components related to programming concepts and numerical computations

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

edik

-0.18

icode

-0.18

runner

-0.15

vailability

-0.15

à¤Łà¤ķ

-0.15

znik

-0.14

ngr

-0.14

aget

-0.14

 Cotton

-0.14

rung

-0.14

POSITIVE LOGITS

´Ŀ

0.17

.NaN

0.16

.toFixed

0.16

mma

0.15

.parseFloat

0.15

phant

0.15

OrNull

0.14

icha

0.14

ics

0.14

Activations Density 0.176%

components related to programming concepts and numerical computations

No Comments

No Known Activations