INDEX

Explanations

references to championship titles and competitions in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

age

-0.18

er

-0.16

iga

-0.15

rah

-0.15

ACC

-0.15

zion

-0.14

ç½®

-0.14

ington

-0.14

tings

-0.14

qi

-0.14

POSITIVE LOGITS

orus

0.20

-winning

0.19

aires

0.18

.scalablytyped

0.17

SHIP

0.17

nable

0.17

reuse

0.17

IMARY

0.17

.gdx

0.16

aire

0.16

Activations Density 0.013%

references to championship titles and competitions in various contexts

No Comments

No Known Activations