INDEX

Explanations

abbreviations and numerical sequences

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

apt

-0.16

-lang

-0.15

enth

-0.15

raman

-0.14

ft

-0.14

.Expressions

-0.14

rack

-0.13

avel

-0.13

inst

-0.13

ouch

-0.13

POSITIVE LOGITS

ALAR

0.17

.requireNonNull

0.16

Ìī

0.15

andro

0.15

ledge

0.15

ecies

0.15

?family

0.14

HeaderCode

0.14

GBK

0.14

hape

0.14

Activations Density 0.899%

abbreviations and numerical sequences

No Comments

No Known Activations