INDEX

Explanations

negative phrases that imply a lack of completion or fairness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

adera

-0.16

Zi

-0.14

unch

-0.14

ouch

-0.14

 lowes

-0.14

bih

-0.14

.spring

-0.14

ë

-0.14

 Narrative

-0.13

Ø§Ø¨Ø©

-0.13

POSITIVE LOGITS

 drain

0.18

itarian

0.15

cie

0.15

Touches

0.15

velt

0.15

Ð·Ð°Ð½Ð¾

0.14

ihan

0.14

ÏĩÎ·

0.14

chine

0.14

 èĩªåĬ¨çĶŁæĪĲ

0.14

Activations Density 0.295%

negative phrases that imply a lack of completion or fairness

No Comments

No Known Activations