INDEX

Explanations

terms and phrases that denote value or worth

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

-0.18

hlen

-0.17

.LA

-0.15

pla

-0.15

ionic

-0.14

shape

-0.14

aye

-0.14

fitness

-0.14

CustomAttributes

-0.14

_requires

-0.14

POSITIVE LOGITS

ness

0.18

 precious

0.16

nes

0.16

 Ø¬Ø¯Ø§

0.16

illy

0.15

 ones

0.15

ly

0.15

/core

0.14

LY

0.14

few

0.14

Activations Density 0.067%

terms and phrases that denote value or worth

No Comments

No Known Activations