INDEX

Explanations

terms related to additional or extra resources or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

 ever

-0.16

oshi

-0.15

_MACHINE

-0.15

_partitions

-0.15

à¥įà¤¤à¤°

-0.14

aju

-0.13

slu

-0.13

opal

-0.13

 absol

-0.13

ntag

-0.13

POSITIVE LOGITS

-than

0.17

bonus

0.16

 than

0.16

 bonus

0.16

 niÅ¼

0.15

Bonus

0.15

iator

0.15

itional

0.15

ologne

0.15

ulti

0.15

Activations Density 0.113%

terms related to additional or extra resources or actions

No Comments

No Known Activations