INDEX

Explanations

programming-related keywords and structure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

ofilm

-0.17

//{{

-0.16

èĩ

-0.16

SizeMode

-0.16

 Ð²Ð²

-0.15

Mailer

-0.15

ÐµÑģÑı

-0.15

ÑģÑĤÑĮ

-0.15

alled

-0.14

emouth

-0.14

POSITIVE LOGITS

est

0.17

 excess

0.15

 Goat

0.15

 contra

0.15

 struct

0.14

ques

0.14

mat

0.14

uss

0.14

æ¨¹

0.14

Activations Density 0.028%

programming-related keywords and structure

No Comments

No Known Activations