INDEX

Explanations

phrases and concepts related to debate and opposition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

morph

-0.18

 morph

-0.17

 Guil

-0.16

 Morph

-0.16

ayi

-0.15

 Forget

-0.14

jian

-0.14

 aren

-0.14

isn

-0.14

Forget

-0.13

POSITIVE LOGITS

 milit

0.20

 bespoke

0.19

 dispose

0.18

 stamps

0.18

 invest

0.18

 rebounds

0.18

vit

0.17

 incl

0.17

 rebound

0.17

 forest

0.17

Activations Density 0.218%

phrases and concepts related to debate and opposition

No Comments

No Known Activations