INDEX

Explanations

instances of legal or judicial language and associated actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 ÑĤÐ¾Ð»

-0.16

 tomorrow

-0.16

zens

-0.14

 Ð·Ð°Ð²ÑĤÑĢÐ°

-0.14

},"

-0.13

Ø´Ø¯

-0.13

Ymd

-0.13

ctions

-0.13

 Chiáº¿n

-0.13

stantiate

-0.13

POSITIVE LOGITS

bid

0.22

 multi

0.21

 rare

0.19

 series

0.17

 lengthy

0.17

 nearly

0.17

 widely

0.16

 taxpayer

0.16

 extensive

0.15

 elaborate

0.15

Activations Density 0.218%

instances of legal or judicial language and associated actions

No Comments

No Known Activations