INDEX

Explanations

mentions of panels in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

es

-0.18

emark

-0.17

afone

-0.17

emp

-0.17

ening

-0.17

esin

-0.16

eil

-0.16

onent

-0.15

annies

-0.15

Ú¯Ø§Ø±

-0.15

POSITIVE LOGITS

led

0.30

ists

0.30

ing

0.24

ize

0.23

ayout

0.22

ist

0.21

ized

0.21

icious

0.20

 discussion

0.20

lica

0.19

Activations Density 0.024%

mentions of panels in various contexts

No Comments

No Known Activations