INDEX

Explanations

the word "ent," indicating a focus on entertainment-related content

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

èŃ

-0.16

ling

-0.16

aisal

-0.15

 Kont

-0.14

anth

-0.13

aller

-0.13

anity

-0.13

à¸µà¸¢à¸£

-0.13

Trade

-0.13

 Shel

-0.13

POSITIVE LOGITS

 Settlement

0.16

isos

0.15

ÑģÑĤÑĢÐ¾Ð¹

0.15

WithMany

0.14

ernes

0.14

hiba

0.14

GS

0.14

nap

0.14

holder

0.14

èĨ

0.14

Activations Density 0.000%

the word "ent," indicating a focus on entertainment-related content

No Comments

No Known Activations