INDEX

Explanations

expressions of announcement or positive news

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

 åħ

-0.15

tern

-0.14

 Everett

-0.14

oui

-0.14

bj

-0.14

heel

-0.14

 fetch

-0.14

zu

-0.13

POSITIVE LOGITS

ann

0.25

 announce

0.24

Ann

0.21

 announces

0.21

 report

0.21

announce

0.21

Ann

0.20

 annon

0.20

 annonce

0.20

 announced

0.19

Activations Density 0.028%

expressions of announcement or positive news

No Comments

No Known Activations