INDEX

Explanations

numerals and numerical information relevant to statistics or data points

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

udent

-0.15

eus

-0.14

dera

-0.14

uth

-0.14

vek

-0.14

semb

-0.13

ibrary

-0.13

£

-0.13

ucchini

-0.13

zig

-0.13

POSITIVE LOGITS

ATAR

0.15

usta

0.14

rien

0.14

inte

0.14

gst

0.14

Ã¶m

0.13

 Boot

0.13

ppard

0.13

asaki

0.13

 among

0.13

Activations Density 0.067%

numerals and numerical information relevant to statistics or data points

No Comments

No Known Activations