INDEX

Explanations

terms related to derivation and origin

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

aise

-0.17

maker

-0.17

izer

-0.17

åįĵ

-0.17

tk

-0.16

-0.15

Beg

-0.15

run

-0.15

ize

-0.15

eller

-0.14

POSITIVE LOGITS

 dara

0.21

-from

0.21

 from

0.20

FromClass

0.19

derived

0.18

fts

0.17

	from

0.17

 dari

0.16

 FROM

0.16

IGNAL

0.15

Activations Density 0.015%

terms related to derivation and origin

No Comments

No Known Activations