INDEX

Explanations

lines of code or configuration settings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

etting

-0.15

agner

-0.15

 Strauss

-0.14

abela

-0.14

ision

-0.14

 successes

-0.13

 densities

-0.13

 vibr

-0.13

 spread

-0.13

POSITIVE LOGITS

elu

0.19

ajes

0.16

hape

0.15

este

0.15

ulo

0.14

 Salem

0.14

odu

0.14

ptest

0.14

Til

0.14

eyin

0.13

Activations Density 0.017%

lines of code or configuration settings

No Comments

No Known Activations