INDEX

Explanations

punctuation marks and exclamations

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

lsru

-0.16

 Puzzle

-0.14

ola

-0.13

 Ingen

-0.13

Reb

-0.13

 Roth

-0.13

Ð°Ð¼Ð¿

-0.12

isas

-0.12

ho

-0.12

 arrang

-0.12

POSITIVE LOGITS

Ð»Ð¸ÑĨ

0.15

cÃ©

0.14

_ASYNC

0.14

ÏĦÎ¿Î¼Î±

0.14

ØŃØ§Ø¯

0.14

timeofday

0.14

okit

0.14

aterno

0.13

ãĥ¼ãĥĬ

0.13

é³

0.13

Activations Density 0.265%

punctuation marks and exclamations

No Comments

No Known Activations