INDEX

Explanations

mentions of individuals, particularly names or handles associated with social media or public figures

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

ropa

-0.16

site

-0.15

geme

-0.14

 Bare

-0.14

MMdd

-0.14

VID

-0.14

erdale

-0.14

 Hague

-0.14

 SPORT

-0.14

_ajax

-0.13

POSITIVE LOGITS

bidden

0.18

antan

0.16

nesc

0.15

 TaÅŁ

0.15

builtin

0.15

ADDE

0.14

Ã±a

0.14

ï¸

0.14

Ð½Ð¾Ðº

0.14

alth

0.14

Activations Density 0.073%

mentions of individuals, particularly names or handles associated with social media or public figures

No Comments

No Known Activations