INDEX

Explanations

terms related to "pop" culture and pop-related phenomena

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

ÙĨ

-0.16

ncia

-0.15

ossal

-0.14

 parallel

-0.14

fty

-0.14

quette

-0.14

ekil

-0.14

Ã±o

-0.14

ervas

-0.14

opyright

-0.14

POSITIVE LOGITS

aram

0.16

.twitter

0.15

boarding

0.15

 indeb

0.14

endum

0.14

uÃ©

0.14

LIKELY

0.14

ateria

0.14

charged

0.13

Activations Density 0.043%

terms related to "pop" culture and pop-related phenomena

No Comments

No Known Activations