INDEX

Explanations

words that indicate actions or roles within discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.15.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

ienen

-0.17

nda

-0.16

owell

-0.15

ÙĦÙĤ

-0.15

-0.14

asley

-0.14

utut

-0.14

Ð²Ð¸Ñī

-0.13

ãĥĥãĥī

-0.13

utive

-0.13

POSITIVE LOGITS

ETO

0.15

tee

0.15

ãĤ´ãĥª

0.15

enstein

0.14

 Rolls

0.14

 Ernest

0.14

 Joint

0.14

anager

0.13

ascar

0.13

 jointly

0.13

Activations Density 0.004%

words that indicate actions or roles within discussions

No Comments

No Known Activations