INDEX

Explanations

concepts related to actions and their consequences

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

iese

-0.17

iez

-0.16

oga

-0.16

ijd

-0.16

alus

-0.16

.entrySet

-0.16

ovah

-0.16

ziej

-0.15

olk

-0.15

ypass

-0.15

POSITIVE LOGITS

inae

0.15

ÑĢÑıÐ´

0.14

monster

0.14

Ã´m

0.14

 sabotage

0.13

 promise

0.13

XHR

0.13

quia

0.13

 Steel

0.13

Activations Density 0.002%

concepts related to actions and their consequences

No Comments

No Known Activations

concepts related to actions and their consequences

No Comments

No Known Activations