INDEX

Explanations

phrases related to health and safety organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

uhn

-0.17

.mb

-0.14

.Clamp

-0.14

 Clamp

-0.14

 Redistributions

-0.14

OUNDS

-0.14

åį±

-0.13

 Grat

-0.13

akan

-0.13

Äĩ

-0.13

POSITIVE LOGITS

CDC

0.28

 Surveillance

0.25

CDC

0.25

 surveillance

0.24

_MM

0.21

MM

0.20

 surve

0.20

 provisional

0.20

cdc

0.20

 Vital

0.20

Activations Density 0.053%

phrases related to health and safety organizations

No Comments

No Known Activations