INDEX

Explanations

temporal phrases and references to future events or sessions

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

anan

-0.17

mlink

-0.16

esel

-0.15

itud

-0.15

loth

-0.14

 Tests

-0.14

ektor

-0.14

à¤¿à¤¯à¤¤

-0.13

 Casinos

-0.13

emean

-0.13

POSITIVE LOGITS

 busy

0.23

 legislative

0.22

 cycle

0.21

busy

0.20

 earnings

0.20

 fall

0.19

 reporting

0.18

 crunch

0.18

 prime

0.17

 school

0.17

Activations Density 0.200%

temporal phrases and references to future events or sessions

No Comments

No Known Activations