INDEX

Explanations

concepts related to depth, exploration, and relationships, particularly in physical and metaphorical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

-0.15

ensburg

-0.15

à¥ĭà¤·

-0.14

á»Ńa

-0.14

airo

-0.14

lag

-0.14

INY

-0.13

ekl

-0.13

POSITIVE LOGITS

 deep

1.22

deep

1.10

 deeper

1.01

 Deep

1.01

 depth

0.99

Deep

0.99

æ·±

0.98

 deepest

0.98

depth

0.91

 depths

0.91

Activations Density 0.223%

concepts related to depth, exploration, and relationships, particularly in physical and metaphorical contexts

No Comments

No Known Activations