INDEX

Explanations

concepts related to hypocrisy and inconsistent beliefs

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

ÙĦØ§Øª

-0.15

 Bien

-0.15

 blas

-0.15

Ã©ric

-0.14

iek

-0.14

uid

-0.14

iu

-0.14

cyn

-0.14

iment

-0.13

ester

-0.13

POSITIVE LOGITS

 straw

0.19

 defenses

0.18

 Straw

0.17

 defenders

0.17

skyt

0.16

 attacks

0.16

oje

0.16

wap

0.15

 defend

0.15

 pole

0.15

Activations Density 0.840%

concepts related to hypocrisy and inconsistent beliefs

No Comments

No Known Activations