INDEX

Explanations

references to inclusivity or the concept of "anyone."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 3-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L3R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.3.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

á»©a

-0.16

zel

-0.14

smarty

-0.14

ibe

-0.14

ively

-0.14

IBE

-0.14

igel

-0.14

itan

-0.14

aw

-0.13

iple

-0.13

POSITIVE LOGITS

 else

0.25

onymous

0.17

_else

0.17

else

0.16

Else

0.15

ãģķãĤĵãģ®

0.15

imetype

0.15

ØªÙĤ

0.15

imeType

0.15

orton

0.15

Activations Density 0.014%

references to inclusivity or the concept of "anyone."

No Comments

No Known Activations