INDEX

Explanations

instances of high numerical values or parameters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

£½

-0.18

oppable

-0.15

ynes

-0.15

erm

-0.15

amedi

-0.14

erp

-0.14

ieren

-0.14

EL

-0.14

@qq

-0.14

POSITIVE LOGITS

rost

0.15

lington

0.15

ROTO

0.15

YLES

0.14

ØºØ§ÙĨ

0.14

mere

0.14

uluk

0.14

ycz

0.14

_firestore

0.14

Activations Density 0.005%

instances of high numerical values or parameters

No Comments

No Known Activations