INDEX

Explanations

negative expressions or phrases, particularly those implying absence or lack

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

ively

-0.18

ulary

-0.15

ej

-0.15

oley

-0.15

sWith

-0.15

yum

-0.15

ression

-0.15

mtree

-0.14

rella

-0.14

avou

-0.14

POSITIVE LOGITS

theless

0.36

-ending

0.25

rr

0.19

-ever

0.17

withstanding

0.17

onta

0.17

ocity

0.16

olution

0.16

itz

0.16

emiah

0.16

Activations Density 0.036%

negative expressions or phrases, particularly those implying absence or lack

No Comments

No Known Activations