INDEX

Explanations

instances of numerical values or counts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

rep

-0.15

zen

-0.15

 Scholar

-0.15

 Unsure

-0.14

 fool

-0.14

iddy

-0.14

uppe

-0.14

.medium

-0.13

yster

-0.13

 Fool

-0.13

POSITIVE LOGITS

iare

0.17

otel

0.16

iez

0.16

_READONLY

0.15

Î¾Î¹

0.15

ffield

0.15

ãĤ¿ãĥ¼

0.14

addtogroup

0.14

iero

0.14

_EOL

0.14

Activations Density 0.000%

instances of numerical values or counts

No Comments

No Known Activations