INDEX

Explanations

the phrase "next" indicating transitions or changes in topics

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

ubi

-0.17

ÐºÐ¾ÑģÑĤÑĮ

-0.15

avery

-0.15

uel

-0.14

è¸

-0.14

brit

-0.14

jt

-0.14

æĢ§çļĦ

-0.13

ounces

-0.13

POSITIVE LOGITS

ADED

0.16

pard

0.15

ATAL

0.14

_deinit

0.14

ICH

0.14

etur

0.14

SCAN

0.13

Ð¾Ð»Ð°Ð¹

0.13

 konkrÃ©t

0.13

dds

0.13

Activations Density 0.002%

the phrase "next" indicating transitions or changes in topics

No Comments

No Known Activations