INDEX

Explanations

references to teams and matchups in sports contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

LECT

-0.15

 HÃ¼k

-0.14

_SCR

-0.14

eos

-0.14

ëĿ¼ëıĦ

-0.14

okia

-0.13

adera

-0.13

baar

-0.13

ãĥ©ãĤ¹

-0.13

ancock

-0.13

POSITIVE LOGITS

-U

0.17

chor

0.16

erna

0.15

vern

0.15

athi

0.14

err

0.14

bek

0.14

ÑĢÐµÐ»

0.14

elle

0.14

avor

0.14

Activations Density 0.075%

references to teams and matchups in sports contexts

No Comments

No Known Activations