INDEX

Explanations

references to legal or accusatory actions

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

New Auto-Interp

Configuration

fnlp/Llama3_1-8B-Base-LXM-8x/Llama3_1-8B-Base-L22M-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.hook_mlp_out

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

ift

-0.16

IFT

-0.15

ilst

-0.14

ekl

-0.14

emo

-0.14

ì¤ĳ

-0.14

ateria

-0.14

ATAR

-0.14

mue

-0.14

urnal

-0.14

POSITIVE LOGITS

 Strauss

0.17

enos

0.17

andin

0.15

Ø§ÛĮØ²

0.15

anders

0.15

itzer

0.15

esa

0.14

å£«

0.14

umont

0.14

 rare

0.14

Activations Density 0.098%

references to legal or accusatory actions

No Comments

No Known Activations