INDEX

Explanations

references to specific seasons, episodes, and stages of shows or competitions

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

ÅĽli

-0.14

cki

-0.14

ght

-0.14

 Bett

-0.14

 Glad

-0.14

abal

-0.14

ca

-0.14

ONEY

-0.14

ighth

-0.13

à¸´à¸¡à¸ŀ

-0.13

POSITIVE LOGITS

(s

0.17

iyas

0.15

rupa

0.14

yll

0.14

ï»¿↵↵

0.14

-long

0.14

edl

0.14

esin

0.14

isos

0.13

illy

0.13

Activations Density 0.051%

references to specific seasons, episodes, and stages of shows or competitions

No Comments

No Known Activations