INDEX

Explanations

negative or conflicting interactions and conditions

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

anny

-0.16

cela

-0.15

anning

-0.14

gles

-0.14

prm

-0.14

ÑģÐ»

-0.14

izoph

-0.13

 Essex

-0.13

amd

-0.13

ÛĮÚ©

-0.13

POSITIVE LOGITS

ed

1.91

edBy

1.06

edb

0.89

edn

0.85

edl

0.85

ED

0.73

edm

0.73

edir

0.67

edata

0.65

edd

0.62

Activations Density 0.411%

negative or conflicting interactions and conditions

No Comments

No Known Activations