INDEX

Explanations

specific names and titles associated with military units, organizations, or notable individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ÅĤa

-0.16

ijo

-0.15

TL

-0.15

pirit

-0.14

izen

-0.13

ibly

-0.13

foundation

-0.13

izard

-0.13

POSITIVE LOGITS

0.16

did

0.15

 will

0.15

rypton

0.14

oog

0.13

Å¡ov

0.13

 Crane

0.13

ondo

0.13

aset

0.13

has

0.13

Activations Density 0.605%

specific names and titles associated with military units, organizations, or notable individuals

No Comments

No Known Activations