INDEX

Explanations

significant instances of negation or emphasis in relation to opinion and argumentation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

OWER

-0.15

lest

-0.14

ije

-0.14

Does

-0.13

ODO

-0.13

enci

-0.13

REFERENCES

-0.13

 ÙĨÙĩ

-0.13

ISCO

-0.13

Ala

-0.13

POSITIVE LOGITS

æĺ¯æĪĳ

0.42

is

0.40

 æĺ¯

0.36

 adalah

0.35

æĺ¯

0.34

 lÃł

0.33

 merupakan

0.31

 ÑıÐ²Ð»ÑıÐµÑĤÑģÑı

0.31

 ÎµÎ¯Î½Î±Î¹

0.28

ãģ®ãģĮ

0.27

Activations Density 0.331%

significant instances of negation or emphasis in relation to opinion and argumentation

No Comments

No Known Activations