INDEX

Explanations

code snippets or programming elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ico

-0.14

 kenn

-0.14

 Venue

-0.14

ÙĪÙĨØ¯

-0.14

å¿Ĺ

-0.14

 organising

-0.14

 manuscripts

-0.14

pis

-0.14

 organizing

-0.14

cod

-0.13

POSITIVE LOGITS

 guest

0.19

hyp

0.18

acker

0.17

guest

0.17

eph

0.17

evi

0.16

 Magnum

0.16

neider

0.16

vm

0.16

VIP

0.16

Activations Density 0.000%

code snippets or programming elements

No Comments

No Known Activations