INDEX

Explanations

programming code and data manipulation elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

eton

-0.15

afone

-0.14

isser

-0.14

imoto

-0.14

dit

-0.14

night

-0.13

 station

-0.13

awah

-0.13

emales

-0.13

Deferred

-0.13

POSITIVE LOGITS

 print

0.19

print

0.16

atem

0.15

Ta

0.15

 amat

0.15

iyim

0.14

Ta

0.14

 inval

0.14

_library

0.14

Activations Density 0.031%

programming code and data manipulation elements

No Comments

No Known Activations