INDEX

Explanations

references to specific concepts or themes of importance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

imbus

-0.15

_OBJC

-0.14

etes

-0.14

borg

-0.14

raf

-0.13

ieres

-0.13

ãĥĨãĥ«

-0.13

iez

-0.13

endon

-0.12

/*č↵

-0.12

POSITIVE LOGITS

anÃ©

0.16

opic

0.16

urally

0.15

ologically

0.15

spiel

0.14

ije

0.14

anyl

0.14

hydr

0.14

opia

0.14

SELL

0.14

Activations Density 0.574%

references to specific concepts or themes of importance

No Comments

No Known Activations