INDEX

Explanations

various contexts in which questions are asked

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ummer

-0.16

-0.15

aira

-0.15

 porad

-0.15

 band

-0.15

alk

-0.15

fed

-0.15

alker

-0.15

adows

-0.15

acid

-0.14

POSITIVE LOGITS

 asked

0.15

_ask

0.14

 questions

0.14

ÑģÑĤÐ¸Ð½

0.14

zes

0.14

_rw

0.13

asz

0.13

 Fragen

0.13

 asks

0.13

gle

0.13

Activations Density 0.176%

various contexts in which questions are asked

No Comments

No Known Activations