INDEX

Explanations

words with special characters or specific grammatical features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ãĥĸãĥª

-0.16

jÃŃ

-0.16

 ê²ĥìľ¼ë¡ľ

-0.15

izard

-0.15

ansom

-0.14

rvine

-0.14

uÃ©

-0.14

[d

-0.14

izza

-0.14

-extra

-0.14

POSITIVE LOGITS

 mind

0.25

az

0.24

 Mind

0.21

ann

0.18

Az

0.17

meg

0.17

mind

0.17

ez

0.17

OTH

0.17

 recip

0.16

Activations Density 0.000%

words with special characters or specific grammatical features

No Comments

No Known Activations