INDEX

Explanations

phrases that include the word "to," particularly in the context of admissions or actions related to wrongdoing

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

Ð·Ñı

-0.15

entine

-0.15

 Orleans

-0.15

EMA

-0.15

ìĽ¨

-0.14

ost

-0.14

ãĥŃãĥ³

-0.14

ÑĢÐµÐ¼ÐµÐ½Ð½Ð¾

-0.14

ens

-0.14

WXYZ

-0.13

POSITIVE LOGITS

&P

0.16

ays

0.16

imonials

0.15

AYS

0.15

 ordin

0.15

olib

0.15

esy

0.14

/goto

0.14

eyi

0.14

Activations Density 0.006%

phrases that include the word "to," particularly in the context of admissions or actions related to wrongdoing

No Comments

No Known Activations