INDEX

Explanations

occurrences of specific letters or symbols, particularly in non-English scripts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

olar

-0.18

zik

-0.15

ym

-0.15

omer

-0.15

uch

-0.15

 Carp

-0.14

ilo

-0.14

warts

-0.14

iz

-0.14

UM

-0.13

POSITIVE LOGITS

Ã¢m

0.21

ãĥ©ãĤ¯

0.17

rame

0.17

Slash

0.17

vise

0.17

reten

0.16

 Slee

0.16

(forKey

0.16

asu

0.16

anik

0.16

Activations Density 0.002%

occurrences of specific letters or symbols, particularly in non-English scripts

No Comments

No Known Activations