INDEX

Explanations

terms related to workplace environments and conditions

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

dir

-0.17

igo

-0.16

uro

-0.16

esch

-0.15

ib

-0.15

way

-0.15

Äĥm

-0.14

ÏĥÎ¹

-0.14

adero

-0.14

åĳĺ

-0.14

POSITIVE LOGITS

 dynamics

0.18

/shop

0.17

 democracy

0.17

rips

0.16

illez

0.16

okino

0.16

vais

0.16

wide

0.16

enha

0.16

 culture

0.15

Activations Density 0.029%

terms related to workplace environments and conditions

No Comments

No Known Activations