INDEX

Explanations

terms related to health, safety, and consumer advocacy regarding products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 gotten

-0.19

nab

-0.17

 Aside

-0.17

Aside

-0.16

 mechanic

-0.16

 behaviors

-0.16

portion

-0.16

 Regardless

-0.15

 unfavorable

-0.15

inbox

-0.15

POSITIVE LOGITS

 satisf

0.17

 modem

0.15

Locker

0.15

 Alec

0.15

tement

0.14

 COMPUTER

0.14

åĳĬ

0.14

DIY

0.14

EMU

0.14

 imaginative

0.14

Activations Density 1.759%

terms related to health, safety, and consumer advocacy regarding products

No Comments

No Known Activations