INDEX

Explanations

references to research and exploration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

irie

-0.16

 Kann

-0.15

å°İ

-0.15

ehr

-0.15

ulling

-0.15

loy

-0.14

ITE

-0.14

essed

-0.14

endir

-0.14

 Williamson

-0.14

POSITIVE LOGITS

ero

0.17

0.14

near

0.14

 Outputs

0.14

 research

0.14

icz

0.14

 near

0.14

idget

0.14

ianne

0.13

DEPTH

0.13

Activations Density 0.002%

references to research and exploration

No Comments

No Known Activations