INDEX

Explanations

references to characters or plot elements related to adventure or action themes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Fucking

-0.18

 fucking

-0.18

 fucks

-0.18

 fucked

-0.17

mamak

-0.15

PLICIT

-0.15

 fuck

-0.15

 FUCK

-0.15

 cazzo

-0.15

ä½

-0.15

POSITIVE LOGITS

 mainland

0.15

 Cartesian

0.15

ashion

0.15

irus

0.14

 Fashion

0.14

 invent

0.14

spy

0.14

 villains

0.14

 inventions

0.14

 Secondary

0.14

Activations Density 0.079%

references to characters or plot elements related to adventure or action themes

No Comments

No Known Activations