INDEX

Explanations

numeric or coded references within a document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

ãĥ³ãĤ°

-0.16

ibar

-0.15

AGIC

-0.14

dds

-0.14

estr

-0.14

bett

-0.14

Stride

-0.14

enko

-0.14

uder

-0.14

Î³Ïī

-0.14

POSITIVE LOGITS

FC

0.18

CF

0.17

EA

0.16

FE

0.16

_consts

0.16

fe

0.15

ad

0.15

ee

0.15

FF

0.15

afe

0.15

Activations Density 0.055%

numeric or coded references within a document

No Comments

No Known Activations