INDEX

Explanations

the repeated use of the word "of."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

à¤¾à¤¹

-0.16

pun

-0.16

fleet

-0.15

 Tender

-0.15

uffs

-0.15

adelphia

-0.15

antry

-0.15

uchen

-0.15

uran

-0.15

pun

-0.15

POSITIVE LOGITS

0.16

icing

0.15

ÑĩÑĥ

0.15

ugins

0.15

atoms

0.14

 ÑĤÑĢÐ¾

0.14

vendors

0.14

 swear

0.14

igon

0.14

/error

0.14

Activations Density 0.024%

the repeated use of the word "of."

No Comments

No Known Activations