INDEX

Explanations

proper names, specifically focusing on individuals and characters

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

Oy

-0.16

ieux

-0.16

 material

-0.16

 Dexter

-0.15

imir

-0.15

OA

-0.15

ick

-0.14

 Material

-0.14

ÙĨØ¯Ø±

-0.14

ira

-0.14

POSITIVE LOGITS

eneg

0.16

usa

0.16

ãĤ¹ãĤ¯

0.15

.Logic

0.14

enville

0.14

ÑĨÐµÐ½ÑĤ

0.14

lÃ¼

0.14

 à¤ĭ

0.14

ëĵĿ

0.14

mployee

0.13

Activations Density 0.029%

proper names, specifically focusing on individuals and characters

No Comments

No Known Activations