INDEX

Explanations

terms related to destructive actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

zan

-0.18

ekil

-0.16

-0.15

olis

-0.15

ÃŃÅ¾

-0.15

eo

-0.14

ook

-0.14

oyo

-0.14

Ã¼m

-0.13

ebi

-0.13

POSITIVE LOGITS

responseData

0.15

aine

0.15

 Cousins

0.15

ingly

0.15

ADX

0.15

 envelop

0.15

pike

0.14

èİ·

0.14

appName

0.14

Ð¾ÑĩÐµÐº

0.14

Activations Density 0.007%

terms related to destructive actions or processes

No Comments

No Known Activations