INDEX

Explanations

phrases indicating processes or actions being performed

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

PLICIT

-0.15

 translateY

-0.15

ksam

-0.15

ä¼¼çļĦ

-0.15

asurer

-0.14

andes

-0.14

gnore

-0.14

oretical

-0.14

 disappe

-0.14

trying

-0.14

POSITIVE LOGITS

 means

0.60

 virtue

0.48

way

0.44

means

0.42

 Means

0.39

-products

0.38

gone

0.36

Means

0.34

 dint

0.34

products

0.34

Activations Density 0.318%

phrases indicating processes or actions being performed

No Comments

No Known Activations