INDEX

Explanations

references to award shows and their hosts

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

ornings

-0.14

agus

-0.14

_masks

-0.14

 Ripple

-0.14

rael

-0.13

onas

-0.13

 Pipes

-0.13

ole

-0.13

ICA

-0.13

POSITIVE LOGITS

plen

0.16

½Ķ

0.14

lien

0.14

ufe

0.14

(ConfigurationManager

0.13

ÑĩÑĥ

0.13

anon

0.13

jmu

0.13

jmp

0.13

-wide

0.13

Activations Density 0.026%

references to award shows and their hosts

No Comments

No Known Activations