INDEX

Explanations

statements that express comparisons of quality and performance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

 excessively

-0.17

 heavier

-0.16

 REALLY

-0.15

agraph

-0.15

ulan

-0.15

 extra

-0.15

ooky

-0.15

vÃŃce

-0.15

UTTON

-0.15

yses

-0.14

POSITIVE LOGITS

 much

0.64

far

0.60

much

0.59

 Much

0.57

Much

0.55

 MUCH

0.52

far

0.50

FAR

0.45

Far

0.43

Far

0.42

Activations Density 0.322%

statements that express comparisons of quality and performance

No Comments

No Known Activations