INDEX

Explanations

terms associated with structural or categorical organization

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

ombo

-0.17

 Pron

-0.16

adin

-0.15

ingles

-0.14

pron

-0.14

ninger

-0.14

à¹Ĥà¸Ļ

-0.14

emem

-0.13

Ð»Ð¾Ð½

-0.13

lon

-0.13

POSITIVE LOGITS

OKIE

0.16

erte

0.15

imple

0.15

ÏĥÏĥ

0.15

0.14

NTAX

0.14

 meis

0.14

ieux

0.14

oint

0.14

EMENT

0.14

Activations Density 0.002%

terms associated with structural or categorical organization

No Comments

No Known Activations