INDEX

Explanations

quotation marks and their associated wording

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

mpar

-0.15

ÐĲÑĢÑħÑĸÐ²

-0.15

EMPL

-0.14

empl

-0.14

 remar

-0.13

 âĢŀ

-0.12

“

-0.12

itial

-0.12

ãģŁãģı

-0.12

ANNOT

-0.12

POSITIVE LOGITS

Oh

0.27

oh

0.27

yeah

0.27

Yeah

0.26

ouch

0.25

Hey

0.24

hey

0.24

ugh

0.24

oops

0.24

Activations Density 0.125%

quotation marks and their associated wording

No Comments

No Known Activations