INDEX

Explanations

words and phrases related to media production and evaluation processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

 flashback

-0.13

 Ways

-0.13

 victim

-0.13

Ð»ÑıÐ²

-0.13

 overall

-0.13

ilet

-0.13

 elites

-0.13

 premises

-0.13

allon

-0.12

.Constants

-0.12

POSITIVE LOGITS

 aspect

0.40

 option

0.35

 feature

0.34

 section

0.34

 portion

0.33

 category

0.31

 phenomenon

0.29

 clause

0.28

aspect

0.28

 mechanism

0.28

Activations Density 0.011%

words and phrases related to media production and evaluation processes

No Comments

No Known Activations