INDEX

Explanations

phrases indicating something that has been previously established or recognized

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

ALLY

-0.15

ãĥ³ãĤ°ãĥ«

-0.15

 chap

-0.15

inch

-0.15

finder

-0.15

isco

-0.14

ally

-0.14

yal

-0.14

 korun

-0.14

ABLE

-0.14

POSITIVE LOGITS

egal

0.17

 Already

0.17

 already

0.16

already

0.16

existing

0.15

-existing

0.15

Already

0.15

 existing

0.15

conto

0.15

çŃĴ

0.14

Activations Density 0.033%

phrases indicating something that has been previously established or recognized

No Comments

No Known Activations