INDEX

Explanations

negative values in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

.hs

-0.16

 Seks

-0.14

POR

-0.14

iffin

-0.14

bred

-0.14

_native

-0.14

omore

-0.14

é³´

-0.13

imary

-0.13

jar

-0.13

POSITIVE LOGITS

anca

0.20

ertz

0.16

 Vaugh

0.15

ULL

0.14

 Prospect

0.14

datatable

0.14

oined

0.14

ulk

0.14

ait

0.14

arrant

0.14

Activations Density 0.021%

negative values in various contexts

No Comments

No Known Activations