INDEX

Explanations

adjectives and adverbs that convey intensity or nuance in descriptions

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

olumn

-0.17

airs

-0.17

optic

-0.16

erk

-0.16

ibur

-0.15

ÌĢ

-0.15

reamble

-0.15

velte

-0.15

iciel

-0.15

dÃ¼r

-0.15

POSITIVE LOGITS

ly

1.74

LY

1.05

ÑģÑı

0.65

lys

0.60

äºİ

0.60

lya

0.50

liness

0.48

lyn

0.44

ness

0.42

ãģ«

0.42

Activations Density 0.292%

adjectives and adverbs that convey intensity or nuance in descriptions

No Comments

No Known Activations