INDEX

Explanations

expressions of excitement and positivity related to racing or competition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

ury

-0.16

anka

-0.14

thur

-0.14

libs

-0.14

anlar

-0.14

tel

-0.14

abis

-0.14

 cutting

-0.14

lett

-0.14

 participant

-0.13

POSITIVE LOGITS

 pole

0.21

 poles

0.21

 podium

0.18

ikk

0.17

 ance

0.17

imin

0.16

bane

0.15

-*

0.14

 Pole

0.14

Pod

0.14

Activations Density 0.035%

expressions of excitement and positivity related to racing or competition

No Comments

No Known Activations