INDEX

Explanations

headings and sections in a structured document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

abi

-0.15

>{!!

-0.15

-0.14

PubMed

-0.13

 seedu

-0.13

 recept

-0.13

bjerg

-0.13

erus

-0.13

 volum

-0.13

ereco

-0.13

POSITIVE LOGITS

encoding

0.18

 Seznam

0.18

coding

0.17

olley

0.16

 coding

0.16

-*-

0.16

%%↵

0.15

ako

0.15

 Lista

0.15

Activations Density 0.005%

headings and sections in a structured document

No Comments

No Known Activations