INDEX

Explanations

code structure and logical operations involving object hierarchy and data types

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ives

-0.14

ãĥĥãĤ·ãĥ¥

-0.14

 outgoing

-0.14

 uninterrupted

-0.14

abee

-0.14

acas

-0.13

rio

-0.13

nar

-0.13

 magg

-0.13

POSITIVE LOGITS

 empty

0.36

 None

0.35

 nothing

0.34

 blank

0.33

 Empty

0.31

 null

0.31

 Nothing

0.31

-empty

0.30

 none

0.30

None

0.30

Activations Density 0.405%

code structure and logical operations involving object hierarchy and data types

No Comments

No Known Activations