INDEX

Explanations

specific objects and quantities related to items or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

thren

-0.18

ucci

-0.15

.bukkit

-0.14

iales

-0.14

ioxide

-0.14

 åī

-0.14

ÐµÐ¼Ð°ÑĤÐ¸

-0.14

muz

-0.13

'autres

-0.13

reon

-0.13

POSITIVE LOGITS

ister

0.16

erez

0.16

osp

0.15

ISTER

0.15

atile

0.15

 Abrams

0.15

sta

0.14

stem

0.14

 Sherman

0.14

sted

0.13

Activations Density 0.086%

specific objects and quantities related to items or actions

No Comments

No Known Activations