INDEX

Explanations

terms related to novelty and the early stages of development

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

atan

-0.15

ÙħØ¬

-0.14

oslav

-0.14

ash

-0.14

pei

-0.13

Ash

-0.13

 door

-0.13

tica

-0.13

thren

-0.13

tas

-0.13

POSITIVE LOGITS

RITE

0.14

abase

0.14

ï»¿#

0.14

ipes

0.14

upo

0.14

ÑĢÑıÐ´Ñĥ

0.14

orex

0.14

ãĤ«ãĥĨ

0.14

annon

0.14

/AFP

0.14

Activations Density 0.187%

terms related to novelty and the early stages of development

No Comments

No Known Activations