INDEX

Explanations

mentions of experiments and experimental practices

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

_subplot

-0.16

emer

-0.15

atan

-0.15

omi

-0.15

ovan

-0.15

pheric

-0.15

ibi

-0.15

ough

-0.14

ogh

-0.14

raphics

-0.14

POSITIVE LOGITS

ERSHEY

0.19

ers

0.15

aly

0.15

avicon

0.14

.jsoup

0.14

chor

0.14

ishly

0.14

itag

0.14

ìĤ¼

0.13

abant

0.13

Activations Density 0.030%

mentions of experiments and experimental practices

No Comments

No Known Activations