INDEX

Explanations

references to government officials, particularly ministers, and their ministries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

-0.17

intosh

-0.16

eer

-0.15

leigh

-0.15

iness

-0.15

scribe

-0.15

eza

-0.15

NESS

-0.14

agan

-0.14

POSITIVE LOGITS

ial

0.28

IAL

0.20

isters

0.19

/max

0.18

ialis

0.18

iales

0.17

ially

0.17

istry

0.16

istrar

0.16

arend

0.16

Activations Density 0.025%

references to government officials, particularly ministers, and their ministries

No Comments

No Known Activations