INDEX

Explanations

adjectives describing quality and condition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

paque

-0.16

Ã¼lÃ¼k

-0.16

ugu

-0.15

alam

-0.14

Ã¥l

-0.14

olang

-0.14

eward

-0.14

ä¾Ľ

-0.13

eton

-0.13

POSITIVE LOGITS

 enough

0.17

 thing

0.16

 testament

0.15

way

0.15

 phenomena

0.14

Ð²ÐµÑī

0.14

 phenomenon

0.14

thing

0.13

TRIES

0.13

ower

0.13

Activations Density 0.223%

adjectives describing quality and condition

No Comments

No Known Activations