INDEX

Explanations

terms related to manipulation or deceitful behavior, especially in social or transactional contexts

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

acias

-0.17

ician

-0.16

á»ĭch

-0.15

æ¦

-0.15

 Linden

-0.15

asis

-0.14

ABCDEFGHIJKLMNOP

-0.14

é¡ĺãģĦ

-0.14

(æĹ¥

-0.14

avin

-0.14

POSITIVE LOGITS

ulation

0.60

ular

0.55

ulated

0.54

ulate

0.53

ulating

0.51

ula

0.50

ulations

0.50

ulator

0.48

ulates

0.47

ulators

0.45

Activations Density 0.038%

terms related to manipulation or deceitful behavior, especially in social or transactional contexts

No Comments

No Known Activations