INDEX

Explanations

references to the concept of "first things."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

ent

-0.15

YST

-0.15

iped

-0.14

ANJI

-0.14

StringUtil

-0.14

chio

-0.14

umba

-0.13

ãĥķãĥĪ

-0.13

QN

-0.13

POSITIVE LOGITS

 thing

0.19

inch

0.16

 Marvin

0.15

Swe

0.15

thing

0.15

aid

0.15

 coisa

0.14

ãĥ©ãĤ¯

0.14

 Humph

0.14

och

0.14

Activations Density 0.034%

references to the concept of "first things."

No Comments

No Known Activations