INDEX

Explanations

code structures and programming constructs

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

SelectionMode

-0.16

ÙĨÛĮÙĨ

-0.15

 Dual

-0.15

pace

-0.14

.walk

-0.14

rum

-0.14

QRS

-0.14

 Broken

-0.13

ikal

-0.13

 Boundary

-0.13

POSITIVE LOGITS

 singleton

0.16

singleton

0.16

ovol

0.15

ve

0.15

/cgi

0.15

oplevel

0.15

 Habit

0.15

dere

0.15

0.14

phinx

0.14

Activations Density 0.071%

code structures and programming constructs

No Comments

No Known Activations