INDEX

Explanations

references to proper nouns, especially names and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

GenerationStrategy

-0.15

NetMessage

-0.14

//{{

-0.14

jedn

-0.14

eyle

-0.14

$MESS

-0.14

mÃ®

-0.13

{.

-0.13

/WebAPI

-0.13

Å¾ila

-0.13

POSITIVE LOGITS

 (“

0.13

 èĩªåĬ¨çĶŁæĪĲ

0.12

↵

0.11

 painstaking

0.11

’

0.11

“

0.10

åºľ

0.10

/Area

0.10

oucher

0.10

Activations Density 5.612%

references to proper nouns, especially names and titles

No Comments

No Known Activations