INDEX

Explanations

references to data tables and columns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

kola

-0.16

anvas

-0.16

rim

-0.16

eldon

-0.16

dos

-0.16

ycl

-0.15

lash

-0.15

lashes

-0.15

slashes

-0.15

sembl

-0.15

POSITIVE LOGITS

ar

0.29

arity

0.25

ists

0.22

aire

0.21

aires

0.21

wise

0.20

ophon

0.18

-wise

0.18

Defs

0.17

atura

0.17

Activations Density 0.023%

references to data tables and columns

No Comments

No Known Activations