INDEX

Explanations

terms associated with sending or receiving messages

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

 Mach

-0.17

 Mold

-0.15

 Minds

-0.15

 Marilyn

-0.15

 Moor

-0.15

 Machinery

-0.15

olis

-0.15

 Maul

-0.14

mute

-0.14

 Mund

-0.14

POSITIVE LOGITS

 message

0.98

 Message

0.82

message

0.82

 mess

0.81

 messages

0.81

-message

0.76

msg

0.75

Message

0.74

 Mess

0.73

_message

0.73

Activations Density 0.275%

terms associated with sending or receiving messages

No Comments

No Known Activations