INDEX

Explanations

references to historical or mythological elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

anship

-0.14

ãĥ³ãĥĪ

-0.14

swick

-0.14

odi

-0.14

_SAMPL

-0.13

.presentation

-0.13

ISCO

-0.13

auga

-0.13

ouns

-0.13

ampling

-0.13

POSITIVE LOGITS

 Atlantis

0.20

atl

0.17

ATL

0.17

 Plato

0.16

Atl

0.16

 Pill

0.15

Atl

0.15

 atlas

0.15

 kred

0.15

onaut

0.15

Activations Density 0.000%

references to historical or mythological elements

No Comments

No Known Activations