INDEX

Explanations

blocks of code and implementation details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

Äĩe

-0.17

acha

-0.16

anes

-0.15

ide

-0.15

ark

-0.14

unt

-0.14

als

-0.14

et

-0.14

ort

-0.13

POSITIVE LOGITS

Tau

0.15

Ø³Ø§Øª

0.15

 defaultManager

0.15

DIR

0.14

irectory

0.14

indow

0.14

ENO

0.14

ssf

0.14

ignum

0.14

SQ

0.14

Activations Density 0.060%

blocks of code and implementation details

No Comments

No Known Activations