INDEX

Explanations

expressions of preference or emphasis in speech

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

oran

-0.17

ilor

-0.16

velt

-0.16

ales

-0.14

amos

-0.14

atories

-0.14

apur

-0.14

uvre

-0.14

gress

-0.13

POSITIVE LOGITS

anch

0.15

rif

0.15

apiro

0.15

 Minority

0.14

 teng

0.14

ÐµÐ½ÐºÐ¾

0.14

 modest

0.13

æ·

0.13

 WikiLeaks

0.13

SAT

0.13

Activations Density 0.002%

expressions of preference or emphasis in speech

No Comments

No Known Activations