INDEX

Explanations

code-related terms and functions, particularly those involving rectangles and graphical elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

.scalablytyped

-0.17

ston

-0.14

ÑĩÐ¸ÑģÐ»

-0.14

æ³Ĭ

-0.14

isas

-0.14

-0.13

Ø§ÙĨÙĪ

-0.13

 environ

-0.13

eniz

-0.13

MATRIX

-0.12

POSITIVE LOGITS

:CGRect

0.15

arem

0.15

 bounds

0.15

irk

0.14

Ð´Ð¾Ð²

0.14

iline

0.14

ned

0.14

illed

0.14

GT

0.14

 Nest

0.14

Activations Density 0.058%

code-related terms and functions, particularly those involving rectangles and graphical elements

No Comments

No Known Activations