INDEX

Explanations

instances of ranked lists or numerical representations

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

reesome

-0.16

rose

-0.15

Ã¼

-0.14

baru

-0.13

ä¼

-0.13

_hello

-0.13

orra

-0.13

 Ø®ÙĪÙĨ

-0.13

icides

-0.13

alive

-0.13

POSITIVE LOGITS

opot

0.16

hus

0.14

enga

0.14

avar

0.14

Fal

0.14

ande

0.14

GED

0.14

&t

0.14

.af

0.13

xd

0.13

Activations Density 0.112%

instances of ranked lists or numerical representations

No Comments

No Known Activations