INDEX

Explanations

references to physical parts or features in a technical or mechanical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

dden

-0.16

LEAN

-0.15

Ã¡v

-0.15

conte

-0.14

onna

-0.14

dom

-0.14

rna

-0.14

fully

-0.14

thes

-0.13

-IS

-0.13

POSITIVE LOGITS

itsu

0.16

/end

0.15

 Instantiate

0.15

uzey

0.15

iro

0.14

ekk

0.14

OfWork

0.14

Å¡k

0.14

arto

0.14

ãĥ³ãĥ

0.14

Activations Density 0.027%

references to physical parts or features in a technical or mechanical context

No Comments

No Known Activations