INDEX

Explanations

references to specific films or cinematic elements

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

-0.15

iveau

-0.14

Ð»ÑĮÑĤ

-0.14

 hÃ´m

-0.13

odiac

-0.13

oes

-0.13

acer

-0.13

ACP

-0.13

usercontent

-0.13

anst

-0.13

POSITIVE LOGITS

 Spice

0.26

 spice

0.25

 sand

0.19

idar

0.18

 Bene

0.18

_LD

0.18

Arr

0.18

 Imperial

0.17

 sands

0.17

 planet

0.16

Activations Density 0.004%

references to specific films or cinematic elements

No Comments

No Known Activations