INDEX

Explanations

historical figures and their titles associated with various historical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 Sentry

-0.15

abic

-0.15

Ã¼p

-0.14

arp

-0.14

ÐµÐ´ÐµÑĢÐ°

-0.14

tplib

-0.14

ajes

-0.13

MBER

-0.13

QUIT

-0.13

eden

-0.13

POSITIVE LOGITS

 himself

0.16

uhe

0.14

chg

0.13

III

0.13

 Ventures

0.13

ÑĢÐ¾Ð²Ð¸Ñĩ

0.13

Ð°Ð³Ð°ÑĤÐ¾

0.13

_Ex

0.12

 reigning

0.12

ÐµÐ´Ð¸Ð½Ð¸

0.12

Activations Density 0.070%

historical figures and their titles associated with various historical contexts

No Comments

No Known Activations