INDEX

Explanations

names of individuals, particularly surnames, and familial connections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

indle

-0.15

isphere

-0.15

raÃ§

-0.14

 Ð¶Ðµ

-0.13

mut

-0.13

undra

-0.13

leness

-0.13

ilee

-0.13

abaj

-0.13

 Ð¸Ð´ÐµÑĤ

-0.13

POSITIVE LOGITS

Ø§ÙĨÙĪ

0.18

 family

0.16

orum

0.15

ikut

0.14

ominator

0.14

(es

0.14

WithMany

0.14

deaux

0.13

éĺħè¯»æ¬¡æķ°

0.13

 siblings

0.13

Activations Density 0.175%

names of individuals, particularly surnames, and familial connections

No Comments

No Known Activations