INDEX

Explanations

instances of the word "or" and its context in choices and alternatives

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

CALE

-0.17

ertino

-0.16

ĥģ

-0.15

ables

-0.15

Î²Î¬

-0.14

maybe

-0.14

orra

-0.14

egers

-0.14

ku

-0.14

burger

-0.14

POSITIVE LOGITS

anger

0.19

phin

0.15

/bower

0.15

thon

0.14

 Convers

0.14

 convers

0.14

ehir

0.14

alsex

0.13

Fan

0.13

rin

0.13

Activations Density 0.011%

instances of the word "or" and its context in choices and alternatives

No Comments

No Known Activations