INDEX

Explanations

references to individuals and their statements or actions

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

 question

-0.19

 Mention

-0.17

 Frage

-0.16

 reference

-0.16

orman

-0.16

 questions

-0.15

talk

-0.15

 wonder

-0.15

 mention

-0.15

-question

-0.15

POSITIVE LOGITS

 maintained

0.25

 maintain

0.23

arg

0.23

 contend

0.23

 maintains

0.22

 argue

0.22

 argued

0.21

 argues

0.20

Maint

0.19

 maintaining

0.19

Activations Density 0.079%

references to individuals and their statements or actions

No Comments

No Known Activations