INDEX

Explanations

the word "First" indicating the beginning of a list or sequence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

orta

-0.17

Ã©r

-0.14

th

-0.14

kin

-0.14

exp

-0.14

rol

-0.14

èĭ±éĽĦ

-0.14

fff

-0.14

most

-0.14

teenth

-0.13

POSITIVE LOGITS

enberg

0.23

enburg

0.21

nett

0.18

Merit

0.17

 Baptist

0.17

eba

0.17

 Presbyterian

0.17

Responder

0.16

Aid

0.16

_unref

0.16

Activations Density 0.036%

the word "First" indicating the beginning of a list or sequence

No Comments

No Known Activations

the word "First" indicating the beginning of a list or sequence

No Comments

No Known Activations