INDEX

Explanations

instances of the word "in" and variations indicating location or context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 whichever

-0.17

erm

-0.15

 whatever

-0.14

utom

-0.14

 whoever

-0.14

 either

-0.14

 wherever

-0.13

 yourselves

-0.13

ä»¤

-0.13

ERM

-0.12

POSITIVE LOGITS

 remarks

0.32

 comments

0.30

 statements

0.25

 interviews

0.23

comments

0.23

 emailed

0.22

remarks

0.22

 tweets

0.22

 letters

0.21

 emails

0.21

Activations Density 0.089%

instances of the word "in" and variations indicating location or context

No Comments

No Known Activations