INDEX

Explanations

concepts related to priorities and prioritization

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

aklÄ±

-0.16

letter

-0.15

ixo

-0.14

oldur

-0.14

illez

-0.14

-hook

-0.14

jour

-0.14

eco

-0.14

/ion

-0.14

IDb

-0.14

POSITIVE LOGITS

pants

0.16

izr

0.16

itag

0.15

ÑĤÐµÑģÑĮ

0.15

upertino

0.15

sed

0.14

ofile

0.14

entially

0.14

ìĪľ

0.14

ìĤ¬íķŃ

0.14

Activations Density 0.011%

concepts related to priorities and prioritization

No Comments

No Known Activations