INDEX

Explanations

references to wrestling and television personalities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

iros

-0.15

bakan

-0.14

éĻ¢

-0.14

acha

-0.14

kontakte

-0.14

aft

-0.14

istent

-0.14

oser

-0.14

afari

-0.14

POSITIVE LOGITS

 host

0.47

 hosts

0.43

 Host

0.42

 anchor

0.42

Host

0.38

anchor

0.37

hosts

0.36

 anch

0.36

 HOST

0.36

 anchors

0.36

Activations Density 0.169%

references to wrestling and television personalities

No Comments

No Known Activations