INDEX

Explanations

phrases indicating attribution or authorship

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

erce

-0.15

ãĥ¼ãĥł

-0.14

.deg

-0.14

onda

-0.14

ulumi

-0.14

éĤ¦

-0.13

_verbose

-0.13

 equipments

-0.13

 FUNCT

-0.13

 dear

-0.13

POSITIVE LOGITS

 similar

0.15

 various

0.14

imity

0.14

 Conc

0.14

 variety

0.13

&id

0.13

 Lloyd

0.13

 actionTypes

0.13

 Piper

0.13

Activations Density 0.297%

phrases indicating attribution or authorship

No Comments

No Known Activations