INDEX

Explanations

instances of numerical or episode-related information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

ukes

-0.18

berman

-0.15

 Lever

-0.15

pecially

-0.15

iro

-0.14

awn

-0.14

uki

-0.14

ç¯

-0.14

fs

-0.14

tir

-0.13

POSITIVE LOGITS

GPC

0.16

loff

0.14

utdown

0.14

.easing

0.14

ãĥĨãĥ«

0.14

ylÃ¼l

0.14

Ã¼zel

0.14

ableObject

0.14

SetActive

0.13

ransition

0.13

Activations Density 0.001%

instances of numerical or episode-related information

No Comments

No Known Activations