INDEX

Explanations

elements related to emotional manipulation and personal storytelling

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

 they

-0.18

they

-0.15

åºľ

-0.14

wins

-0.14

we

-0.14

eree

-0.14

asm

-0.14

 KromÄĽ

-0.14

 Schwartz

-0.14

ifes

-0.14

POSITIVE LOGITS

 that

0.22

 ÑĩÑĤÐ¾

0.20

 That

0.17

 THAT

0.17

_that

0.17

 Å¼e

0.17

que

0.17

 hogy

0.17

éĤ£ç§į

0.16

 ÑīÐ¾

0.16

Activations Density 0.279%

elements related to emotional manipulation and personal storytelling

No Comments

No Known Activations