INDEX

Explanations

code-related structures and object properties

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eer

-0.17

opy

-0.15

agara

-0.15

apy

-0.14

ari

-0.14

undler

-0.14

enko

-0.14

ortho

-0.14

Ã«n

-0.14

POSITIVE LOGITS

ãĥ¼ãĥĨãĤ£

0.15

ìĹŃ

0.15

emann

0.15

epam

0.15

yne

0.15

 confer

0.15

Aura

0.14

Ø®ØªÙĩ

0.13

_YUV

0.13

ichtet

0.13

Activations Density 0.127%

code-related structures and object properties

No Comments

No Known Activations