INDEX

Explanations

the concept of "two" or "three" as it relates to options, ways, or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

var

-0.14

icht

-0.14

Enc

-0.14

ichten

-0.14

GCC

-0.14

 casting

-0.13

vs

-0.13

ottes

-0.13

POSITIVE LOGITS

.asp

0.17

ltra

0.16

essim

0.16

Detach

0.15

_bases

0.15

Ð»Ð¾Ð¿

0.15

¤ëĭ¤

0.14

ombok

0.14

ahir

0.14

å±Ģ

0.14

Activations Density 0.046%

the concept of "two" or "three" as it relates to options, ways, or categories

No Comments

No Known Activations