INDEX

Explanations

terms related to direction, paths, and future outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

 exchanged

-0.16

afone

-0.15

urovision

-0.15

ÑĢÐµÐ¶

-0.15

BackStack

-0.15

ahy

-0.14

pii

-0.14

 exchange

-0.14

pedia

-0.14

lena

-0.14

POSITIVE LOGITS

rok

0.15

umen

0.14

 Fellow

0.14

 Fell

0.14

sha

0.14

gang

0.14

 Bened

0.14

 gent

0.14

ectors

0.14

 toward

0.14

Activations Density 0.152%

terms related to direction, paths, and future outcomes

No Comments

No Known Activations