INDEX

Explanations

parentheses and their use in categorizing content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

á»Ń

-0.17

plain

-0.17

gard

-0.16

zych

-0.16

inton

-0.15

icha

-0.15

ringe

-0.14

arter

-0.14

ardu

-0.14

POSITIVE LOGITS

ìĶ¨

0.16

leftright

0.14

.dt

0.14

addtogroup

0.14

_ONCE

0.13

 Levine

0.13

_acl

0.13

ÎĨ

0.13

Forgery

0.13

 Kramer

0.13

Activations Density 0.026%

parentheses and their use in categorizing content

No Comments

No Known Activations