INDEX

Explanations

phrases and terms emphasizing visual perception and attention

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

anism

-0.16

ional

-0.16

aire

-0.16

upplier

-0.15

(SIG

-0.15

/install

-0.15

ghi

-0.15

Ð¾Ð²ÑĸÐ´

-0.15

lement

-0.14

eurs

-0.14

POSITIVE LOGITS

ight

0.23

-eye

0.20

-catching

0.19

IGHT

0.17

prints

0.16

lessly

0.15

ockets

0.15

·»

0.15

SED

0.15

Activations Density 0.038%

phrases and terms emphasizing visual perception and attention

No Comments

No Known Activations