INDEX

Explanations

the presence of disclaimers or statements regarding fictional content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

hoot

-0.18

BeforeEach

-0.17

HOOK

-0.16

dbcTemplate

-0.14

hook

-0.14

optgroup

-0.14

bjerg

-0.14

Ð²ÐµÑĤ

-0.13

âĶģâĶģ

-0.13

ãĥ¬ãĥ¼

-0.13

POSITIVE LOGITS

sha

0.16

nard

0.15

venir

0.15

 handicap

0.15

riad

0.15

kos

0.15

endez

0.14

Geh

0.14

 Train

0.14

 service

0.14

Activations Density 0.009%

the presence of disclaimers or statements regarding fictional content

No Comments

No Known Activations