INDEX

Explanations

references to percentages and proportions in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

omm

-0.17

inan

-0.15

scape

-0.15

 Malone

-0.15

urg

-0.15

 Seks

-0.15

zik

-0.14

lo

-0.14

spi

-0.14

roperties

-0.14

POSITIVE LOGITS

(%)

0.22

oft

0.18

ake

0.16

of

0.16

0.15

ighted

0.15

heads

0.15

alte

0.15

hare

0.15

ety

0.14

Activations Density 0.029%

references to percentages and proportions in various contexts

No Comments

No Known Activations