INDEX

Explanations

pronouns and references to people 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 @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

only

-0.18

 also

-0.17

 cannot

-0.16

 only

-0.16

also

-0.16

 likewise

-0.15

oloj

-0.15

Cannot

-0.14

antar

-0.14

 ALSO

-0.14

POSITIVE LOGITS

 exactly

0.33

 actually

0.24

 ultimately

0.22

 might

0.22

actually

0.21

 Exactly

0.21

 vlastnÄĽ

0.20

 Actually

0.20

/how

0.20

Actually

0.19

Activations Density 0.156%

pronouns and references to people in various contexts

No Comments

No Known Activations