INDEX

Explanations

instances of the word "as."

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

ÅĻi

-0.16

.Slf

-0.15

ceae

-0.15

à¸©

-0.14

PackageManager

-0.14

quee

-0.14

ARRANT

-0.14

issance

-0.14

iaux

-0.14

steen

-0.14

POSITIVE LOGITS

irk

0.17

lice

0.16

peria

0.16

inch

0.16

abi

0.15

.sy

0.15

int

0.15

onn

0.15

oka

0.14

Activations Density 0.043%

instances of the word "as."

No Comments

No Known Activations