INDEX

Explanations

references to significant individuals and their actions or circumstances

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

Negative Logits

Propagation

-0.16

vetica

-0.16

|--------------------------------------------------------------------------↵

-0.15

enda

-0.15

apot

-0.14

tur

-0.14

 propagation

-0.14

UnderTest

-0.14

mond

-0.14

SetName

-0.14

POSITIVE LOGITS

è¥²

0.15

etti

0.14

linger

0.14

 bases

0.14

 invol

0.14

Lyn

0.14

 bang

0.14

 Giles

0.13

ILED

0.13

kas

0.13

Activations Density 0.004%

references to significant individuals and their actions or circumstances

No Known Activations