INDEX

Explanations

programming-related syntax and structure, particularly in relation to markup languages and data manipulation

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

iny

-0.15

 circ

-0.15

åł¡

-0.15

elson

-0.14

imore

-0.14

olson

-0.14

 Hundred

-0.13

uster

-0.13

ÑīÑĸ

-0.13

.synthetic

-0.13

POSITIVE LOGITS

adera

0.16

 Dolphin

0.15

ceptar

0.15

orp

0.15

Ã¼c

0.15

Äħd

0.14

'options

0.14

ader

0.14

GP

0.14

ØªØ´

0.13

Activations Density 0.026%

programming-related syntax and structure, particularly in relation to markup languages and data manipulation

No Comments

No Known Activations